, LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative Action/Equal Opportunity institution omens-o1 KINETIC AND PHARMACOLOGICAL CHARACTERIZATION OF MURINE PGH SYNTHASE-l AND PGH SYNTHASE-Z By Elizabeth Anne Meade A DISSERTATION Submitted to Michigan State Universtiy in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physiology 1992 .(f/V / 6w 6}??- ABSTRACT KINETIC AND PHARMACOLOGICAL CHARACTERIZATION OF THE MURINE PGH SYNTHASE-l AND PGH SYNTHASE-2 By Elizabeth Anne Meade We have isolated a mouse cDNA clone for a second isozyme of PGH synthase, the central enzyme in prostaglandin biosynthesis. Expression of this enzyme, PGHSm-Z, and PGHSm-l in COS-1 cells allowed comparison of the kinetic and pharmacological properties of these isozymes. PGHSmu-l and PGHSmu-Z have nearly the same KIn for arachidonic acid, 3.0 and 2.5 pm respectively. PGHSm-Z however, utilizes eicosatrienoic acid and eicosapentaenoic acid more efficiently than PGHSmu-l. The non-steroidal anti- inflarnmatory drugs (NSAIDs) indomethacin, sulindac sulfide and piroxicam preferentially inhibit PGHSm- 1; while ibuprofen, flurbiprofen, meclofenamate and docosahexaenoic acid are equipotent inhibitors of the isozymes. One drug, 6-MNA, the active metabolite of Nabumetone, preferentially inhibits PGHSmu-Z. These initial results suggest that drugs that are completely selective for either isozyme can be found. The most surprising difference in the response of these isozymes to NSAIDs was seen with aspirin. While aspirin inhibits PGHSmu-l oxygenation of arachidonate completely, treatment of PGHSmg-Z with aspirin results in a enzyme which converts arachidonate to 15-hydroxyeicosatetraenoic acid instead of POI-12. The identity of this product was confirmed by gas chromatography/mass spectroscopy. Regulation of expression of PGHSm-l and PGHSm-Z is also different. Using a 3T3 cell system, we found that PGHSm-l was expressed constitutively. PGHSmu-Z, however, was expressed only transiently following serum stimulation. Induction of PGHSmu-Z expression occurred at the level of transcription of the gene. Dexamethasone inhibited serum-induced transcription of the PGHSmu-Z gene and enzyme expression, but had no detectable effect on PGHSm-l expression. Thus, we propose the following model for the role of these isozymes. PGH synthase-l is expressed constitutively in most tissues and can therefore produce prostaglandins in immediate response to hormones. Prostaglandins produced by PGH synthase-l are probably involved in regulation of homeostasis, and in housekeeping functions. PGH synthase-2 is expressed only in stimulated cells. Prostaglandins produced by this enzyme are probably involved in inflammation, mitogenesis, or developmentally regulated processes such as ovulation. It is probable that the anti-inflammatory effects of dexamethasone are in part due to inhibition of PGH synthase-Z expression. Copyright by ELIZABETH ANNE MEADE 1992 ACKNOWLEDGMENTS It would have been impossible for me to complete my graduate school work without the help from a large number of people. My thesis director, Dr. David DeWitt has been extraordinarily patient while I stumbled through learning about molecular biology and cellular work. He has taught me much about research and, by example, the importance of putting in the long hours when necessary to get the job done. Dr. William Smith has also played a major role in teaching me about prostaglandins and how to really _th_igk_ about my project by asking tough questions at the right times. My guidance committee including Dr. Harvey Sparks, Dr. William Spielman, and Dr. Steve Heidemann has been very helpful in steering my research and helping me complete all the necessary requirements to finish graduate school. My time in graduate school would not have been the same without the help of a number of fellow students, post docs and other faculty who supported me both intellectually and emotionally. Although there is not enough time or space to mention them all, I want to highlight just a few. Stacey Kraemer has been both a friend and an instructor, working with me in the first few painful months when I hardly knew what DNA was. The other members of the lab, particularly Marty Regier and Odette Laneuville have been supportive and helpful. A number of friends outside the lab have been very important as well; particularly Bob Cole, Dave LeVier and Patti Tithof who have commiserated with me about the pain and suffering of graduate school and helped me learn how to survive in one piece with my sanity intact. Without all of these people, graduate school would have been a much less enjoyable place. Finally, I couldn’t have gotten through it all without the support of my family. Never have they doubted my ability to be smart enough and tough enough to get through this program. Thanks to everybody! TABLE OF CONTENTS LIST OF TABLES ........................................ viii LIST OF FIGURES ....................................... ix LIST OF ABBREVIATIONS ................................. xii CHAPTER 1: LITERATURE REVIEW Introduction ........................................ 1 Physiological roles of prostaglandins ...................... 3 Regulation of PGH synthase expression .................... 5 Are there multiple PGH synthase genes? .................... 12 Non-steroidal anti-inflammatory drugs ..................... 16 CHAPTER 2: IDENTIFICATION, EXPRESSION AND KINETIC CHARACTERIZATION OF PGH SYNTI-IASE-Z Introduction ........................................ 21 Methods .......................................... 22 Results ........................................... 40 Discussion ......................................... 66 CHAPTER 3: PHARMACOLOGICAL CHARACTERIZATION OF PGH SYNTHASE-l AND PGH SYNTHASE-Z Introduction ........................................ 71 Methods .......................................... 73 Results ........................................... 78 Discussion ......................................... 102 CHAPTER 4: SERUM AND GLUCOCORTICOID REGULATION OF GENE TRANSCRIPTION AND EXPRESSION OF PGH SYNTHASE-l AND PGH SYNTHASE-Z Introduction ........................................ 106 Materials and methods ................................ 109 Results ........................................... 1 15 Discussion ......................................... 141 CONCLUSION .......................................... 145 vi BIBLIOGRAPHY ......................................... 15 1 vii Table 1. Table 2. Table 3. Table 4A. Table 4B. Table 5. Table 6. LIST OF TABLES Factors which affect PGH synthase expression in vitro ..... 6 Non-steroidal anti-inflammatory drugs ................ 19 Comparison of PGHSmu-l and PGHSmu-Z .............. 47 V,mm ratios for PGHSmu-l and PGHSmu-Z using different substrates ............................... 62 Vmu values for PGHSm-l and PGHSmu-Z for individual experiments .............................. 62 Two distinct members of the PGH synthase gene family by homology ............................... 67 Inhibition of PGHSmu by non-steroidal anti-inflammatory drugs .................................. 83 viii Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. LIST OF FIGURES Biosynthetic pathway for prostaglandin formation ........ 2 Putative scheme for the regulation of PGH synthase expression ........................ 7 Structures of non-steroidal anti-inflammatory drugs ....... 17 Plasmids for transient expression of PGH synthase in COS-1 cells ............................... 25 Construction of expression plasmids containing the coding region of PGHSm-Z ................... 27 Sequences of the ovine 5’UTR amplified by PCR for ligation with PGHSmu-l and PGHSm-Z ........... 30 Construction of expression plasmids containing the coding region of PGHSmu-l ................... 33 Construction of PGHSOV expression plasmids ............ 36 Northern blot analysis of mouse 3T3 cell mRNA hybridized with a 1.7 kb EcoRI fragment from the sheep PGH synthase CDNA at high (50% formamide) and low (30% formamide) stringencies . . 41 Restriction analysis of PGHSmu-l and PGHSmu-Z showing the size and location of the EcoRI (R1) restriction fragments ........................ 43 Comparison of the amino acid sequences of PGHSmu-l and PGHSmu-Z ............................ 45 PGHSm-Z expression plasmids and related cyclooxygenase and peroxidase activities ..................... 49 ix Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. PGHSm-l expression plasmids and related cyclooxygenase and peroxidase activities ..................... 49 PGst expression plasmids and related cyclooxgenase and peroxidase activities ..................... 49 Oxygen electrode traces of cyclooxygenase activity of microsomes from cos-1 cells transfected with the PGH synthase expression plasmids .............. 52 Determination of the Km for PGHSmu-l and PGHSmu-Z for arachidonic acid ........................ 56 Structures of fatty acid substrates tested for cyclooxygenase activity with PGHSmu-l and PGHSm-Z ........... 59 Comparison of the Vmu for PGHSmu-l and PGHSmu-Z using arachidonic acid (AA) or eicosatrienoic acid (ETA) as the substrate ........................... 61 Determination of the Km for PGHSmu-l and PGHSmu-Z for eicosatrienoic acid (ETA) ................. 63 Comparison of the V“ for PGHSm-l and PGHSm-Z using arachidonic acid (AA) or eicosapentaenoic acid (EPA) as the substrate ........................... 65 Inhibition of PGHSmu cyclooxygenase activity by flurbiprofen .............................. 79 Inhibition of PGHSm cyclooxygenase activity by ibuprofen ............................... 80 Inhibition of PGHSm cyclooxygenase activity by meclofenamic acid ......................... 81 Inhibition of PGHSmu cyclooxygenase activity by docosahexaenoic acid (22:6) .................. 82 Inhibition of PGHSmu cyclooxygenase activity by indomethacin ............................. 85 Inhibition of POI-IS,“ cyclooxygenase activity by sulindac sulfide ........................... 86 Figure 27. Figure 28. Figure 29. Figure 30. Figure 31. Figure 32. Figure 33. Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Figure 39. Figure 40. Inhibition of PGHS,“ cyclooxygenase activity by piroxicam ............................... 87 Inhibition of PGHSm cyclooxygenase activity by 6-MNA . . . 88 Time course for inhibition of PGHSmu cyclooxygenase activities by aspirin ........................ 89 Product formation by PGHSmu-l and PGHSmu-Z treated with various NSAIDs ....................... 91 Mass spectrum for the non-prostaglandin product synthesized by PGHSm-Z in the presence of aspirin . . 93 Time course of aspirin inhibition of prostaglandin formation ............................... 95 The effects of varying concentration of aspirin on the ability of the PGHSmu isozymes to synthesize prostaglandins ............................ 98 Northern blot analysis of serum induction of PGH synthase-1 and PGH synthase-2 mRNAs in NIH 3T3 cells ............................ 116 Transcription of the PGH synthase-1 and -2 genes as determined by nuclear run-off assays ............ 121 Dose-response curve for dexamethasone inhibition of the serum-induced PGH synthase-2 mRNA expression . . .126 Effect of time of addition of dexamethasone on inhibition of serum—induction of PGHSm-Z mRNA .......... 130 RU486 reverses dexamethasone inhibition: TPA is more sensitive to dexamethasone inhibition ............ 132 Western blot analysis of PGH synthase expression in serum- and serum plus dexamethasone-treated 3T3 cells ............................... 136 Model for regulation of prostaglandin synthesis by PGH synthase-1 and PGH synthase-2 ............ 148 xi 3’UTR 5’UTR 6-MNA lS-HETE 7L 1114’ AA BAEC COX dex. EPA ETA fibro. GMCSF hCG HUVEC IDso IL-l K... LH nGRE N SAID PDGF PG PGHSm- l PGHSm-Z PGst PO follicles POX TLC TPA V ABBREVIATIONS 3’ untranslated region 5’ untranslated region 6-methoxy-2-naphthyl acetic acid 15-hydroxy-eicosatetraenoic acid lambda phage macmphage arachidonic acid bovine aortic endothelial cells cyclooxygenase dexamethasone eicosapentaenoic acid eicosatrienoic acid fibroblast granulocyte macrophage colony stimulating factor human chorionic gonadotropin human umbilical vein endothelial cells inhibitory dose 50 interleukin- 1 Michaelis-Menton constant luteinizing hormone negative glucocorticoid regulatory element non-steroidal anti-inflammatory drug platelet derived growth factor prostaglandin murine prostaglandin synthase, isozyme 1 murine prostaglandin synthase, isozyme 2 ovine prostaglandin synthase preovulatory follicles peroxidase thin layer chromatography 12-O-tetradecanoyl 13-acetate maximum velocity of a reaction xii CHAPTER 1 LITERATURE REVIEW Introduction Prostaglandins are autocoids which participate in regulation of many physiologically important processes such as vascular homeostasis, renal water reabsorption, ovulation, parturition and inflammation (1-4). Prostaglandins are not stored within cells but rather are produced in response to stimulation by a wide variety of hormones, growth factors, and proteases (5, 6). Through mechanisms that remain unclear, these agents interact with cell surface receptors to activate phospholipases, which, in turn, release arachidonic acid from phospholipid stores (Figure l). Arachidonic acid can then be converted to PGH2 by the enzyme PGH synthase (cyclooxygenase). PGH synthase has two distinct enzymatic activities: a bis-oxygenase activity which incorporates two molecules of oxygen into arachidonic acid at carbon positions 9- and 11-, and 15- to yield PGGZ, and a peroxidase activity which reduces PGG2 to yield PGHz. PGH2 is a common intermediate in the synthesis of all prostaglandins, however, final conversion of PGH2 to biologically aetive prostaglandins is tissue-specific and depends on the expression of various PGH2 isomerases and reductases. For example, endothelial cells contain PGI2 synthase and produce prostacyclin, platelets contain thromboxane synthase and synthesize TxAz, and ovaries contain PGHz-PGFZG reductase and produce PGFm. W8 ACTIVATION OF 3 HO PHOSPHOLIPASES ' 2e '’6“: W on P662 “0‘ CCAAW MN NO 6.1 PGFZ. r; J: ecu. MEMBRANE Figure l. Biosynthetic pathway for prostaglandin formation. 3 Physiological roles of prostaglandins In the last decade one of the most studied roles of prostaglandins has been their regulation of vascular homeostasis. Endothelial cells produce prostacyclin (PGI,), a vasodilator and inhibitor of platelet aggregation (7). Platelets produce thromboxane (TxAz), a vasoconstrictor and platelet aggregator (8). Together, the two work in opposition to each other within the context of a much more complex biochemical regulatory system to maintain vascular integrity and blood flow in regions of injury or infection. The relative importance of regulating the balance of synthesis of P012 and TxA2 for maintaining vascular homeostasis has been illustrated by studies of large groups of people who ingest aspirin regularly (2, 9). Inhibition of vascular prostaglandin synthesis by aspirin, which, at appropriate doses can preferentially reduce TxA2 synthesis, is thought to reduce platelet aggregation and thrombus formation (10). The tipping of the balance in favor of prostacyclin synthesis appears to significantly reduce the risk of myocardial infarction and the complications that ensue (2, 9). Another important physiological role of prostaglandins is the control of blood flow, renin release and water reabsorption in the kidney. PGE, and prostacyclin produced within the kidney, cause vasodilation and increase renal blood flow. These two prostaglandins also attenuate the effects of vasoconstrictive hormones such as angiotensin II and norepinephrine (11). PGan, produced in the kidney in response to bradykinin, and TxAQ, produced by platelets, are potent vasoconstrictors that reduce renal blood flow. PGE2 stimulates the production of renin, and is involved in regulating AVP-induced water reabsorption in the collecting tubule; important processes for regulation of whole body salt and water balance. While it has been suggested that prostaglandins play a direct role 4 in hypertension, the relationship has never been directly proven. It is likely that they are produced only in response to hypertension and are not involved in the development of hypertension (11). One potential pathophysiological situation in which prostaglandins play a role is inflammation. Prostaglandins, particularly PGE, produced by macrophages and fibroblasts at the site of injury or infection, stimulate extravasation of fluid from blood vessels and enhance the pain response (12, 13). Moncada et al. (13) proposed that the role of prostaglandins in inflammation is to potentiate the effects of primary mediators of inflammation, such as bradykinin or histamine. Inhibition of prostaglandin formation by either steroidal or non-steroidal anti-inflammatory drugs thus does not prevent inflammation, but does limit the extent of the inflammatory response. A final example of the role of prostaglandins is in regulation of ovulation, an event that has been compared to inflammation (14). In estrus cycle animals, including the sheep and rat, there is a surge of LH just prior to ovulation. In these animals it has been shown that the LH surge leads to the production of large quantities of prostaglandins which are required for the rupture of the follicle and release of the ovum(a) (15, 16). 5 Regulation of PGH synthase expression Production of prostaglandins is regulated at a number of different levels (Figure 1). The first level of regulation is at the activity of phospholipases which release arachidonic acid from membrane stores. Interaction with many different types of stimuli with target cells results in the activation of phospholipases and the subsequent release of arachidonic acid (6). In many cells and tissues liberation of arachidonic acid leads to an immediate short burst of prostaglandin production via constitutively expressed PGH synthase. Release of arachidonic acid, however, is often not sufficient to allow sustained synthesis of prostaglandins as PGH synthase can be rapidly auto-inactivated catalytically (17, 18). Continued synthesis of prostaglandins thus requires synthesis of new PGH synthase. Many of the factors which stimulate arachidonic acid release including interleukins (19), growth factors (20, 21) and phorbol esters (22) also stimulate new prostaglandin synthesis (See Table 1, Figure 2). Two types of regulation of PGH synthase expression occur. The first type, typically observed in cultured fibroblasts and endothelial cells, involves regulation of steady-state levels of prostaglandin synthesis. In these cells, PGH synthase mRNA can increase 2-3 fold and prostaglandin synthesis increases 5- to 20-fold, while there may be small (2-3 fold) or no increase in PGH synthase protein levels. Examples of the second type of regulation are seen in the ovary and in promonocyte lines, such as U937 and HL-60, following differentiation to macrophage-like cells. This type of regulation involves developmentally-induced expression of PGH synthase in cells which previously expressed little or no PGH synthase protein. Prostaglandin synthesis in HL-60 cells increases 50- to 100-fold, and is accompanied by substantial increases in PGH synthase protein levels (up to 15-fold) (23). 6 Table 1. Factors which affect PGH synthase expression in vitro + + + 21, 24-26 mesangial + + + + 27 CEF-l47 fibro. NR NR NR 4» 28 C127 fibro. NR + + + 29. 30 IL] SV-T2 313 + + NR NR 31 dermal fibro. + + +" NR 19. 32, 33 HUVECS + NR + + 34, 35 monocytes NR NR NR + 30 PMA dermal fibro. + + + NR 32 amnion N1? + NR NR 36 HUVECS + NR + NR 22 HL-60 - + + NR 21. 37 U937 + + + NR 33 3T3 cells NR NR NR + 24-26 LH PO follicles NR + + + 1, 16. 38. 39 progesterone uterus NR + + + 40, 41 EGF amnion NE + NR NR 40, 42, 43 MC3T3-El + + 4» NR 44 bovine smooth muscle cells «0- + NR NR 45 3T3 cells NR NR NR + 26 11,2 HUVECS + NR + NR 46 BAECS + NR + NR 46 Epinephrine MC3T3-EI + + + NR 47 FRTL-S + NR NR NR 48, 49 TGF-B MC3T3-El + + + NR 50 LPS U937 4» NR NR NR 51 monocyte/mo + + NR NR 33, 52. 53 Dexametlmsone monocytes/m¢ — - NR NR 33. 52, 53 U937 - - - NR 33 dermal fibro. - - -‘ -‘ 32, 33 3T3 cells - NR - - 25, 54 C127 fibro. NR - - - 30 a. NE, has no effect ; NR, was not reported; (4»), increases synthesis, aetivity or expression; (-), decreases synthesis, activity or expression. b. Increased ”S-rnethionine labeling, but protein levels were not determined. c. Decreased ”S-methionine labeling, but protein levels were not determined. d. Decreased in vitro translatable mRNA levels, total PGH synthase mRNA levels were not determined Figure 2. Putative scheme for the regulation of PGH synthase expression. FaCtors that increase prostaglandin synthesis by increasing PGH synthase expression may do so by increasing transcription of the gene(s). Regulation of PGH synthase mRNA levels may also be at the level of mRNA stability. PGH synthase protein may feed back and inhibit its own formation. Glucocorticoids may also inhibit the production of PGH synthase either at the transcriptional or translational level. LH Putative PUG: Posflve lL-1 Regulatory PMA Agents EGF CAMP LPS PGH SYNTHASE ——>PGH SYNTHASE __> PGH SYNTHASE GENE transcription mRNA translation PROTEIN (degradation) k (7). . / Glucocortuconds re Putative Synthesis Negative Regulatory ' Elements Inactivation + proteoiysis FIGURE 2 9 The first type of regulation, steady-state regulation of PGH synthase expression, was first observed in mouse 3T3 cells in experiments by Habenicht et al. (21). 3T3 cells (a mouse fibroblast cell line) were allowed to become quiescent in medium devoid of growth factors, then were treated with platelet derived growth factor (PDGF). PDGF stimulated a biphasic induction of prostaglandin synthesis; an early acute increase in which prostaglandin synthesis reached a plateau within two minutes, and a second sustained production two hours later to a level more than twice the initial level. The authors found that inhibition of protein synthesis did not alter the initial burst of prostaglandin production but eliminated the second burst, suggesting that new PGH synthase formation was required to allow the continued formation of prostaglandins. Other experiments supported this concept. The authors measured prostaglandin synthesis from exogenously added arachidonic acid and found that the total capacity of the cells to synthesize PGEZ was increased two-fold 4 hours after the addition of PDGF further suggesting that new PGH synthase was produced following stimulation with PDGF. Kinetic studies on microsomes from PDGF-stimulated cells showed a 10-fold increase in the V“ of PGH synthase 16 hours following PDGF stimulation, while there was no change in the K,n (55). The second burst of prostaglandin production was also dependent on the uptake of low density lipoproteins (LDL), as a source of arachidonic acid, from the medium (56). PDGF therefore appears to affect prostaglandin synthesis in 3T3 fibroblasts in three distinct ways: 1) it causes the acute release of arachidonic acid from phospholipid stores for the immediate synthesis of prostaglandins, 2) it stimulates the formation of new PGH synthase enzyme, and 3) it leads to enhanced LDL uptake which provides an additional source of arachidonic acid for prostaglandin synthesis. Studies by 10 a number of other groups have shown this pattern of regulation of PGH synthase expression and prostaglandin production is typical of fibroblasts and endothelial cells in response to a wide range of compounds (See Table 1). The increases in LDL uptake have only been demonstrated in 3T3 cells. The second form of regulation of PGH synthase is best exemplified by events that occur in follicles during ovulation, although it is similar to events that occur in differentiating macrophages (55, 57). Prostaglandins play a crucial role in the rupture of the follicle and release of the ovum at ovulation (14, 15). The mechanism by which prostaglandin production is increased prior to ovulation has been shown to involve dramatic increases in PGH synthase protein (1). Hedin and coworkers used quantitative Western blotting to measure changes in PGH synthase expression during ovulation. They found that PGH synthase protein levels in follicles induced to ovulate with human chorionic gonadotropin (hCG) began to rise by 3 hr following treatment, reached a maximum 15-fold above control levels at 7 hr, then returned to basal levels by 12 hrs. PGH synthase was induced specifically in the granulosa cells of these preovulatory follicles and induction was dependent on doses of hCG that induced ovulation. Surprisingly, the authors found by hybridization to a mouse CDNA, that the 2.7 kb rat PGH synthase mRNA did not increase following hCG treatment in spite of the large changes in protein levels observed (16). Prostaglandin production can be negatively regulated as well as stimulated. Anti- inflammatory glucocorticoids such as dexamethasone have been shown in some cases to decrease stimulated PGH synthase expression (33) (See Figure 2), although glucocorticoids do not appear to inhibit constitutive prostaglandin synthesis. 11 Dexamethasone inhibits the IL-1 stimulated production of prostaglandins in mouse fibroblasts, and in mouse and human monocytes (33, 52). Dexamethasone was also shown to decrease the IL—1 stimulated induction of an in vitro translatable mRN A in both these cells. Surprisingly, while there were changes in a translatable mRNA, these authors presented no data to show changes in PGH synthase mRNA levels, although cDNA probes for a PGH synthase were earlier isolated in this laboratory (33). 12 Are there multiple PGH synthase genes? Because there appear to be two distinct mechanisms for regulating PGH synthase expression, and because of the discrepancies in changes in PGH synthase mRNA levels preceding PGH synthase enzyme induction in follicles, many researchers began to suspect that there might be multiple genes coding for PGH synthase; one constitutively-expressed gene whose enzyme levels were maintained at constant levels, and a second inducible gene expressed only in response to stimuli such as hCG or LH or following differentiation of cells like promonocytes. Another reason to suspect that there might be multiple PGH synthases came from studies of the actions of non-steroidal anti-inflammatory drugs (NSAIDS). The primary mechanism of action of NSAIDS is the inhibition of PGH synthase activity. However, the various classes of these drugs differ in their effects. Different tissues are sensitive to different drugs. Also the potency of drugs differ from cell to cell in their three primary modes of action: anti-inflammatory, anti-pyretic and analgesic. An example of differential tissue-sensitivity to NSAIDS is seen with sulindac. Most NSAIDS must be used cautiously by pe0ple with compromised kidney function because these drugs can accelerate damage or lead to complete kidney failure (58). Sulindac, however, does not inhibit kidney prostaglandin synthesis and is therefore considered safe for those people with compromised kidney function (59). Differences in the potencies of NSAIDS in their modes of action are also common. Acetaminophen, for example, works well as an analgesic and an anti-pyretic but is ineffective as an anti-inflammatory (59). The existence of two (or more) PGH synthases with different tissue distributions and sensitivities to NSAIDS would be one possible explanation for these observations. 13 A final reason to suspect that there might be multiple PGH synthase genes was the disparities seen in the response of various tissues to glucocorticoid regulation. In one set of experiments, mice were pretreated with either endotoxin, dexamethasone or a combination of the two (52). When macrophages were isolated from mice treated with endotoxin, these cells showed an increased ability to synthesize prostaglandins. Dexamethasone inhibited the endotoxin-stimulated increase in synthetic capacity. Dexamethasone, however, did not block basal levels of prostaglandin production in macrophages (52), nor did it block the prostaglandin synthesis in mouse kidney (52). Furthermore, endotoxin treatment did not increase prostaglandin production in mouse kidney. These results suggested that there might be two pools of PGH synthase, one which is induced by endotoxin and whose induction was inhibited by dexamethasone; and a second PGH synthase which was not responsive to either of these agents. Similar results were obtained with human monocytes (53). Taken together these data provide strong circumstantial evidence for the existence of two PGH synthase genes. To address the question of multiple genes in this laboratory, we used Southern blot analysis employing the CDNA for the sheep vesicular gland PGH synthase and its mouse homologue as probes. Two types of Southern blot experiments were performed. The first type involved hybridizing Southern blots of sheep genomic DNA with small restriction fragments from the ovine PGH synthase CDNA (60). In theory, if the restriction fragments used for hybridization were small enough to correspond to single exons, these probes will recognize only single bands on the genomic Southems if a single gene exists and multiple bands if multiple genes exist. We found that one band of genomic DNA hybridized for each restriction we fragment used, suggesting that there was only one gene 14 closely related to the ovine PGH synthase. The second type of Southern blot we performed employed mouse 3T3 cell DNA and took advantage of the fact that we had a complete genomic map for a mouse PGH synthase gene (61). For these experiments, two large pieces of the cDN A, corresponding to the entire coding region of the mouse cDNA were hybridized with Southern blots of restriction digested mouse genomic DNA. Since we knew the complete restriction map of the mouse gene, we knew the number and size of fragments that would hybridize with our mouse cDNA. Only those bands that would be predicted to hybridize hybridized. Thus, we concluded that there were no other PGH synthase genes, at least no genes closely enough related to cross-hybridize. Despite our evidence for a single gene, later work continued to indicate that there were multiple genes. As stated earlier, when Wong et al. (16) initially examined the effects of hCG and LH treatment on isolated rat ovarian follicles, they found that treatment of the whole follicles or the granulosa cells alone with hCG or LH led to an increase in immunologically detectable PGH synthase. Northern blot analysis of the mRNA from these cells surprisingly showed that there was a decrease in the level of a 2.1 kb mRNA for PGH synthase during this period when protein levels increased. One possible explanation was that changes in PGH synthase protein levels did not involve increased transcription of the PGH synthase gene. A second possibility was that there were two enzymes which were both recognized by a single antiserum, and that the mRN A for one of these genes did not hybridize with the other gene. Later work by this same group validated this second theory (62). In a further series of experiments these investigators using antisera from different rabbits, detected a distinct pattern of PGH synthase expression in the rat ovary. The first antiserum (P0869) 15 recognized a 69 kDa PGH synthase expressed in the small antral and preovulatory follicles of the rat ovary. Expression of this PGH synthase are low and constitutive, and did not change following hCG stimulation. This antiserum also reacted with PGH synthase from a number of other rat tissues including uterus, kidney, heart, and adrenal gland. The second antiserum (PGS72) recognized a 72 kDa PGH synthase that is only expressed in the ovary following hCG stimulation. It did not detect the constitutively expressed 69 kDa PGH synthase. These data provide strong evidence for the existence of two different related forms of PGH synthases. A final piece of evidence for a second gene was provided by Rosen er al. (63), who measured prostaglandin production in freshly isolated sheep tracheal epithelial cells and cells which had been cultured for seven days. They found that cultured cells produced much higher levels of prostaglandins than did those cells which were freshly isolated. The authors thought they could detect an increase in immunoprecipitated PGH synthase in the cells cultured for seven days but could not detect a change in PGH synthase message levels (Northern blot analysis). When the Northems were hybridized at lower stringency, a larger band ( 4.0 kb) was evident. This band was more intense in the cultured cells than in the fresh cells. The authors suggested that this message corresponds to an alternate form of PGH synthase which could be responsible for the increased prostaglandin synthesis they saw in tracheal epithelial cells following culture. More recently, we and a number of other investigators (26, 64) have identified cDNAs for the second PGH synthase gene that was predicted above. The isolation and characterization of this second PGH synthase will be the subject of the remaining chapters of this dissertation. 16 Non-steroidal anti-inflammatory drugs Non-steroidal anti-inflammatory drugs (N SAIDs) are among the most widely used drugs in the world. A primary mechanism of action of these drugs is inhibition of the cyclooxygenase activity of PGH synthase. Although aspirin is the most familiar NSAID, there are a wide range of these drugs which have been grouped in families according to their chemical structures (see Figure 3 and Table 2). NSAIDS inhibit PGH synthase in two ways: reversibly and irreversibly. Simple reversible NSAIDS include drugs such as ibuprofen, mefenamic acid, piroxicam, sulindac sulfide, and 6-methoxy-2-naphthyl acetic acid (6-MNA). These drugs are thought to inhibit PGH synthase by competing directly with arachidonic acid for binding to PGH synthase (65). Inhibition by these drugs is not permanent, and removal or dilution of the drug can reverse its effects. NSAIDS which inhibit PGH synthase irreversibly include indomethacin, flurbiprofen and meclofenamate. The mechanism of irreversible inactivation is unknown for these latter N SAIDs, although it is believed to involve high affinity non-covalent interaction of each drug with PGH synthase (6). Aspirin interacts with PGH synthase both reversibly and irreversibly. The reversible interaction is instantaneous and competitive and occurs only at very high inhibitor concentrations (20 mM) and is likely due to the interference of aspirin with arachidonate binding (65). Irreversible inhibition of PGH synthase by aspirin is time- dependent and involves acetylation of an active site serine (66). This residue has been identified as serine 530 in the sheep PGH synthase enzyme, but there are analogous serine residues in the mouse and human proteins. Acetylation of this amino acid interferes with the binding of arachidonic acid so that cyclooxygenase activity cannot occur (66). Replacing the serine residue with another bulky residue by in vitro mutagenesis has been 17 Figure 3. Structures of non-steroidal anti-inflammatory drugs. Structures of NSAIDS used in studies with PGHSm-l and PGHSmu-Z. The compounds are grouped according to the family to which they belong. For a more detailed list of NSAIDS and their families, see Table 2. FLURBIPROFEN F CH \ COOH IECLOFENAHATE CI 1 Mi CH3 Cl SULINDAC SULFIDE CHas cr H | CH2000H F CH3O ASPIRIN coon o I occu3 INDOHETHACIN 18 IBUPROFEN CH; CH COOH PIROXICAH OH _ CONH \ / N /N\ s/ I \ CH3 0 o BollETHOXY-Z-NAPHTHYLACETIC acro (rs-um) 0C coon C - 0 H300 CH3 cnzcooH DOCOSAHEXAENOIC ACID FIGURE 3 19 Table 2. Non-steroidal anti-inflammatory drugs. FAMILY COMPOUND' OXICAMS isoxicam piroxicam SALICYLATES aspirin diflunisal acetamidophenol acetaphenetidin acetaminophen ACETIC ACIDS indomethacin acemetacin tolmetin sulindac sulfide diclofenac zomepirac 6-MNA FENAMATES flufenamic acid niflumic acid meclofenamic acid mefanamic acid PROPIONIC ACIDS naproxen indoprofen ibuprofen flurbiprofen suprofen ketoprofen carprofen PYRAZOLES phenylbutazone oxyphenbutazone ‘ Bolded compounds were tested for inhibition of PGHSm-l and PGHSmu-Z. 20 shown to have the same effect as acetylating the serine residue, providing further evidence that interference with arachidonate binding is the mechanism by which aspirin’s acetylation of this residue interferes with cyclooxygenase activity. As mentioned above, there are differences in the actions of NSAIDS. Although the mechanism of action is the same for almost all NSAIDS, inhibition of PGH synthase, some function better as analgesics than anti-inflammatories (59) or vice versa. There are also tissue-specific actions (or inactions) of particular NSAIDS, as discussed above for sulindac. Identification of a second PGH synthase raises the obvious question of how NSAIDS interact with this new isozyme. This topic will be discussed in Chapter 3. While NSAIDS are some of the most commonly used drugs, they are not without their side effects. The best known side-effect of NSAID therapy is gastrointestinal bleeding which occurs because prostaglandins both stimulate secretion of the protective mucus coating and inhibit acid secretion in the stomach (67). Regular use of high doses of NSAIDS is likely to cause dyspepsia, a common cause of patient noncompliance. Use of NSAIDS for rheumatoid arthritis alone is thought to result in 2600 deaths and 20,000 hospitalizations due to NSAID-associated gastropathy (58). A second serious side-effect of NSAIDS is renal dysfunction. As mentioned above, prostaglandins play a number of roles in normal kidney function including controlling renal blood flow, glomerular filtration and transport of ions across tubular epithelia. NSAIDS can lead to a wide range of renal complications including reversible impairment of glomerular filtration, acute renal failure, hyperkalemia and chronic renal failure (58). It is not surprising that NSAIDS should have serious side—effects because prostaglandins are involved in the regulation of numerous physiological events. CHAPTER 2 IDENTIFICATION, EXPRESSION AND KINETIC CHARACTERIZATION OF PGH SYNTHASE-2 Introduction Although genomic Southern blots done in our laboratory suggested that there was a single gene closely related to PGH synthase (60), data from a number of other laboratories suggested the existence of a second gene (see Chapter 1) In addition, our earlier Northern blot analysis had demonstrated (25) that the Egg; PGH synthase CDNA hybridized with two distinct mouse mRNA species, a 2.7 kb species corresponding to the mouse homologue of the sheep vesicular PGH synthase and a larger 4.5 kb mRNA corresponding to an unknown protein. Thus, hybridization results with the sheep CDNA suggested that there might be two mouse genes for PGH synthase. However, hybridization of the mouse PGH synthase cDNA with a mouse Northern blot showed only a single 2.7 kb hybridizing species, suggesting that hybridization to the second mRNA might have been an artifact peculiar to the sheep CDNA. To resolve this question, we decided to clone a CDNA corresponding to the second, larger mRNA and determine its sequence. ‘21 22 Methods Northern blot afllvsis. Poly A+ RNA for Northern blot analysis was isolated from 3T3 cells by the guanidinium/cesium chloride method (68) followed by oligo dT chromatography. Poly A+ mRNA was electrophoresed on a formaldehyde/agarose gel and transferred to nitrocellulose (68). This Northem blot was hybridized with a 1.6 kb EcoRI fragment of the ovine PGH synthase CDNA (69) which was labeled by random primer synthesis using [a32P]-dCTP (NEN, 800 Ci/mmol) (70). This 1.6 kb EcoRI fragment contains almost the entire coding region of the sheep PGH synthase CDNA. Hybridization was performed at either high stringency (50% formamide, 42°C) or low stringency (30% forrrramide, 42°C) in 5 x SSPE, 1 x Denhardt’s, 0.1% SDS, 1 mg/ml denatured calf thymus DNA and 2 x 10‘5 cpm per ml radiolabeled probe for 18 hours (68). Blots were washed 2 x 15 minutes with 2 x SSC containing 0.1% SDS at room temperature and exposed to XAR-S x-ray film overnight. Screening of the mouse kgtlo libffl A 2.3 kb fragment of the ovine PGH synthase CDNA consisting of the entire coding region and adjacent 5’ and 3’ non-coding regions was radiolabeled as above and used to screen a Agth library made from serum- stimulated Swiss 3T3 cells (66). An initial screening of 1 x 10‘ plaques was performed at low stringency (30% formamide) to identify all clones related to PGH synthase. Positively hybridizing plaques were differentially screened using hybridization solutions containing 30% and 50% formamide, to determine those clones which hybridized well only at lower stringency. Several clones identified in this manner were amplified and phage DNA was isolated (68). 23 Phage purification. k phage were grown on plates at a density of 4 x 10" plaques per 150 mm plate for 6 hours at 37°C. Plates were overlaid with k-diluent (20 mM TrisCl, pH 8.0, 10 mM MgC12) and stored overnight at 4°C. The supematants from the plates were combined and centrifuged for 10 minutes at 10,000 rpm at 4°C. The supernatant was transferred to clean tubes and an equal volume of 20% PEG containing 2 M N aCl was added. Bacteriophage were allowed to precipitate at 4°C for 2 hours and were collected by centrifugation at 9500 rpm for 10 minutes. Phage were resuspended in l-diluent and extracted with CHC13. The aqueous and organic phases were separated by centrifugation for 10 minutes at 3750 rpm, and the aqueous phase was then loaded onto CsCl step gradients in Beckman Ultraclear centrifuge tubes and centrifuged for 2 hours at 30,000 rpm at 20°C (68). Phage DNA (p=l.5), which appear as a grey band in the interface between the CsCl solutions of p=1.3 and p;1.6, were removed with a syringe needle and 1/10 volume 2 M TrisCl, pH 8.5, 0.2 M EDTA and 1 volume formamide were added. The following day 1 volume of H20 and 6 volumes EtOH were added to precipitate the DNA. The phage DNA was next rinsed with 70% EtOH and transferred into clean microfuge tubes. Phage DNA was resuspended in TE and analyzed by restriction digestion with EcoRI endonuclease. EcoRI restriction fragments were subcloned into M13mp19 for sequencing. Sequencing was performed with Sequenase® polymerase according to manufacturer’s protocols. Expression of PGHSm -1 and -2 in a transient expression system. Both the murine PGHSm-l and PGHSm-Z (as the PGH synthase related CDNA was called) cDNAs were subcloned into pSVT7 and pSVL for transient expression in the cos-1 monkey cell line. Our initial reason for wishing to express these cDNA’s was to determine if the protein 24 the PGHSm-Z CDNA coded for exhibited the same cyclooxygenase and peroxidase activities of PGHSm-l. pSVT7 and pSVL are vectors designed for high level transient expression in cos cells. Both vectors contain a polylinker region for insertion of foreign DNA sequences, an SV40 promoter to allow transcription of inserted message, and signal sequences to ensure proper processing of their transcripts (Figure 4). The major difference between the two plasmids is the promoter they contain: pSVT7 uses the SV40 early promoter, while pSVL uses the SV40 late promoter. Cos cells constitutively express wild type SV40 T-antigen and contain all of the factors that are necessary to drive the replication of SV40-origin containing plasmids, such as pSVT7 and pSVL (68). Plasmid copy number can reach as high as 105 copies per cell. Efficient transcription from the plasmids leads to very high expression of the mRNA derived from the cDNA insert, and of the resulting protein. For subcloning of PGHSmu-Z (Figure 5), a 3.26 kb Earl/HindIII fragment was isolated from the lambda clone, Agt10-15.2 (Figure 5A). This fragment begins 6 nucleotides after the translational start site and contains the sequence coding for the last 602 of 604 amino acids of PGHSmu-Z and 0.8 kb of the 3’ untranslated region (UTR). Because Earl digestion eliminates the first two amino acids (6 bp), it was necessary to synthesize an oligonucleotide linker to rebuild the translational start site and the preceding , Kozac translational start sequence. This linker contained EcoRI and Earl cohesive ends to allow the fragment to be ligated into pSVT7 and to the PGHSmu-Z CDNA in the proper orientation (Figure 5B). The HindIII site at the 3’ end of the cDNA was next converted to a SalI site by linker ligation (Figure 5B). This allows the CDNA to be removed from 25 Figure 4. Plasmids for transient expression of PGH synthase in COS-1 cells. pSVT7 and pSVL are vectors designed for the transient expression of foreign cDNAs in cos-1 cells. The vectors contain a polylinker region for insertion of foreign DNA. The SV40 origin of replication (SV40 ori) allows for the replication of the plasmid to high copy number in cos-1 cells. Transcription of the inserted cDN A initiates at the SV40 promoter, the SV40 early promoter in pSVT7 and the SV40 late promoter in pSVL. The vectors contain sequences to ensure proper splicing and polyadenylation of the foreign DNA. 26 SV40 ori Polyllnker .. Sv40 early _ mm promoter Hindill Amp-R "335;, SV40 small T intron SV40 poly A 'amHl 0.90 pSVT7 3.80 Kb SV40 ori Polyiinker SV40 late Xhol Xbal promoter Smal Sacl BamHl ‘ SV40 poly A Sail 0.50 FIGURE 4 27 Figure 5. Construction of expression plasmids containing the coding region of PGHSm-Z. A) The CDNA insert of lgth clone 15.2 containing the PGHSmu-Z cDNA. This cDNA was removed from the vector by digestion with Earl and HindIII restriction endonucleases. B) For ease in subcloning, the HindIII site in the 3’UTR was converted to a SalI site. An oligonucleotide linker cassette was synthesized which contained the translational start site of PGHSmu-Z and a Sall site on the 5’ end. The PGHSmu-Z and oligonucleotide cassette were ligated into pSVT7 to form the plasmid pSV'l‘7-PGHSm-2. C) To shorten the 3’UTR of PGHSmu-Z, the CDNA was digested with EcoRV and SalI linkered. Simultaneously, a sequence containing the ovine 5’UTR (indicated by the slashed bar) was amplified from the ovine PGHS0v cDNA by PCR using oligonucleotides that had EcoRI/Earl ends. The PGHSmu—Z with the shortened 3’UTR and the amplified ovine 5’UTR were ligated into pUC19 to give pUC19-PGHSmu-2A3’UTR. D) The CDNA insert, including the ovine S’UTR was digested out of pUC19 using SalI and ligated into pSVT7 to form pSVT7-PGHSm-2A3’UTR. E) The remainder of the 3’UTR of PGHSmu-Z was removed by digestion with Ach. This restriction site was then SalI linkered and PGHSmu-Z was ligated into both pSVL and pSVT7, forming pSVL-PGHSmu-Z and pSVT7-PGHSmu-21Q3’UTR. 28 wall! A) lambda gtto-PGHSmu-z l l 4‘ lambda gt10-15.2 l I W Hindlll a) pSVT7-PGHSmu-2 1 I ' Sell Elfl ACCI my Sell 0) pUC19-5'ovlne UTR- PGHSmU-2A3'UTR ovine 5' UTR PGHSmu-2 coding region fill/l Ezizizirizltififitizi:3:1:3:1:'-:?:3:?:1:'-'~‘:3zizizifiziziizist'ci:13:11-25:11?:1:?:’13:1:-'43:13:31 : J] pUC19 5009' Earl Accl Sail Sill D) pSVW-PGHSmU-ZAZ‘I'UTR ovine 5‘ UTR PGHSmu-Z coding region [Tl] // ];:;.,;:;:¢:;:;:;1545;113:5555;::;;:-:;:;:;:;:;:f.:;;:§:§.f:§:§:;.;:;:i:§.;;.:;:;:-_:»:§;§;;:L:;l l J pSVT7 f I ' ' 5." En" Acci Sail E) pSVL-PGHSmu-2 and pSVW-PGHSmu-z HQ 3'UTR ovine 5' UTR PGHSrnu-Z coding region [7//// [5:253:,:;:;:;'.;:;:;:;'.-_;;:;:;:-j._~:;:;:;;;;;.;:-_;;:;:I:;:;5:5513:2553;_;.;.;:;.;.:.;:;:;:;. I l Sall Earl Sell pSVL or pSVT7 FIGURE 5 29 the plasmid by a single SalI digestion. Because this initial construct (Figure SB) lacked detectable enzyme activity in transfection experiments, two further modifications were made in the PGHSmu-Z cDNA. First, the 3’ untranslated region (UTR) was shortened by digesting the clone with EcoRV. This restriction site was then blunt ended and SalI linkers were inserted. It was hoped that removal of this portion of the 3’UTR, which eliminated 626 nts and 4 AUUUA sequences--a sequence which decreases mRNA stability (71), would contribute to message stability and thus increase the level of expression of PGHSmu-Z. Next, we added the 5’UTR from the ovine PGH synthase (Figure 5C). We speculated that sequences from the S’UTR of the ovine PGH synthase, which is expressed at very high levels in cos-1 cells, might also increase the efficiency of expression of PGHSmu-Z. A fragment containing the ovine S’UTR was prepared by polymerase chain reaction (PCR). Oligonucleotide primers were designed to allow synthesis of a 105 bp fragment containing 85 bases of the ovine 5’UTR, the first two amino acids of PGHSm-Z; moreover, this fragment contained restriction sites (Sal/Earl) for proper ligation with the PGHSm-Z CDNA (Figure 6A). The plasmid pSVT7-PGHSOV served as the template for the PCR reaction. PCR was performed as described by Sambrook et al. (68). The amplified DNA was run on an agarose gel, and the fragment was isolated using a Geneclean Il® Kit (Bio 101) and subcloned into a pUC19 T-vector (72)(73). The pUCl9-PCR recombinant was sequenced to verify that it contained the desired DNA fragment. Once the sequence was verified, the PCR product was isolated from the pUC19 . vector following EcoRI and Earl digestion using low melting point agarose electrophoresis. The CDNA containing the shortened 3’UTR was cut with Earl and SalI 30 Figure 6. Sequences of the ovine S’UTR amplified by PCR for ligation with PGHSm-l and PGHSm-Z. A) Ovine 5’UTR as constructed for PGHSmu-Z. Two oligonucleotides (in UPPER CASE) were synthesized. The 5’ oligonucleotide contained restriction sites for Xbal and Sall for ease of ligation. The 3’ oligonucleotide contained both the translational start site and an Earl site for ligation with the PGHSmu-Z cDNA. The region between these oligonucleotides was amplified by PCR using the pSVT7-PGHSm, as the template. B) Ovine 5’UTR as constructed for PGHSmu- 1. Two oligonucleotides (in UPPER CASE) were synthesized. The 5’ oligonucleotide was the same as used for amplification of the sequence for PGHSm-Z and contained Xbal and Sall restriction sites for ease of ligation. The 3’ oligonucleotide contained both the translational start site and a BspHI site for ligation with the PGHSmu-l cDNA. The region between these oligonucleotides was amplified by PCR using pSV'l’7-PGHS0v as the template. A) B) 31 Xbal Sall l___l F—'l ATI'CTAGAGTCGACGAGGGGCCGGAGCTCCnggcagagtt agag acgcactccaggagcctg agtcggtctccagcaCGCAACGGC CACCCTGCACCATGCTCTTCCGAGC l_l l__J Earl t/I start site Xba] Sall ATTCTAGAGTCGACGAGGGGCCGGAGCTCCnggcagagtt agag acgcactccaggagcctg agtcggtctccagcaCGCAACGGC CACCCTGCATCATGAGTA i___l Bspl-ll L__i t/l start site FIGURE 6 32 and together with the amplified ovine S’UTR (EcoRI/Sall cut) was ligated into pUC19 cut with EcoRI/Sall (Figure 5C). This resulting ovine 5’UTR-PGHSmu—2 cDNA could be digested from pUC19 as a single 2.0 kb SalI fragment. This fragment was subcloned into the expression plasmid pSVT7 and is referred to as pSVT7-PGHSm-2A3’UTR (Figure 5D). We were also unable to detect enzyme activity with this construct so we further shortened the PGHSmu-Z 3’ end by digestion with Accl to remove the entire 3’UTR, including all AUUUA sequences. A Sall site was again added by linker insertion (68). This final insert was then subcloned into the Sall sites of a second expression vector, pSVL (Figure 5E) as well as into pSVT7. Use of the second vector, pSVL, was to determine whether our initial failures at expression were vector dependent. For subcloning the murine PGH synthase-1 gene PGHSmu-l (Figure 7), a 2.45 kb BspHI/Xbal fragment containing the translational start site, the entire coding region of PGHSm-l and 0.47 kb of 3’UTR was isolated from a full length thlO—PGHSm-l CDNA clone (Figure 7A). BspHI restricts PGHSmu-l very close to the translational start site. By methods similar to those used for PGHSmu-Z, an oligonucleotide cassette containing the ovine 5’UTR was synthesized by PCR. The oligonucleotides used to create the PCR product are shown in Figure 6B. Following synthesis of the 5’UTR for PGHSmu-l it was ligated into pUC19 using the T—vector method. The fragment was then cut out with EcoRI and Bsle and, in a three way ligation, the ovine 5’UTR, PGHSm-l cDNA and pUC19 were joined together (Figure 7B). The entire insert (including the ovine S’UTR) was cleaved from pUCl9-PGHSmu-l with Sall and inserted into the expression vector pSVT7 (Figure 7C). 33 Figure 7. Construction of expression plasmids containing the coding region of PGHSm-l. A) The lgth clone containing the PGHSmu-l cDNA insert. The CDNA was removed from the vector by digestion with BspHI and Xbal. B) For subcloning into pUCl9 the PGHSmu-l sequence was ligated with a fragment containing the ovine 5’UTR (indicated by the slashed bar) which was amplified by PCR with EcoRI/BSpHI ends, to form pUCl9-PGHSmu-l. C) For expression of PGHSmu-l, the PGHSmu-l from pUC19-PGHSm-l was removed from the vector by digestion with Sall. This cDNA insert was then ligated into pSVT7 to form pSVT7-PGHSmu-l. D) The 3’UTR was shortened by digestion of pUCl9-PGHSmu-l with BbsI. The restricted end was blunted and Sall linkers applied. Religation of PGHSmu-l with pUC19 yielded pUC19-PGHSmu- 1A3 ’UTR. E) For expression of the PGHSmu-l with the shortened 3’UTR, the cDNA could be removed from pUC19 by Sall digestion and ligated into either pSVL or pSVT7 to create the plasmids pSVL-PGHSmu-1A3’UTR and pSVT7-PGHSmu-1A3’UTR. 34 Ween 1.9.9.191 A) lambda gt10-PGHSmu-1 PGHSrnu-1 coding region ; t l i i r lambda gt10 BopHI Bbsl Xbal B) pUC19-PGHSmu-1 ovine 5' UTR PGHSrnuci coding region fill/Ill l : g : pUC19 '5le Bele Bbsl Xbal Sell Sell 0) pSVU-PGHSmu-1 OViflO 5' UTR PGHSrnu-1 coding region 577”“: ' i i 4' pSVT7 Sall Bele at... Xbal Sall D) pUC19-PGHSmu-1A3'UTR ovine 5' UTR PGHSmu-1 coding region [vi/1111' l—i pUC19 Sall BeDHI Sall E) pSVL-PGHSmuA3'UTR and pSVT7-PGHSmu-1A3'UTR ovine 5’ UTR PGHSrnu-i coding region ¥////4 l_i pSVL or pSVT7 Sal Hopi-ll Sall FIGURE 7 35 To determine the effects of the 3’UTR on PGHSm-l expression, we shortened the 3’UTR of PGHSmu-l by digesting pUC19-PGHSmu-l with HindIII and BbsI. The ends were blunted, Sall linkered, and reannealed. This digestion removed 595 bases of the PGHSm-l 3’UTR, leaving 46 bases 3’ of the translational stop codon. This construct, called PGHSmu-1A3’UTR (Figure 7D) was isolated from pUC19-PGHSmu-1A3’UTR using Sall digestion and ligated into pSVT7 and pSVL (Figure 7E). Ovine PGH synthase subcloned into pSVT7 was used as a control in transfections (Figure 8A). The entire PGHS0v was cleaved from pSVT7 with Sall and ligated into pSVL (Figure 8B). To determine the effect of the ovine 3’UTR on expression, this region was also truncated by digestion with DraIll and BamHI. The DNA was isolated on NA45, Sall linkered and religated into pSVL to give the vector pSVL-PGHSOVA3’UTR (Figure 8C). Mfection of cos-1 cells. Cos-1 cells were grown to near confluence in DME supplemented with 8% calf serum and 2% fetal calf serum (DME-8+2). The day prior to transfection the cells were cut 1:2 into 100 mm plates. Transfections were performed by the DEAE dextran/chloroquine method (66). Media was removed from the cells and 3 ml DME containing 250 pg DEAE-dextran per ml and 15pg of expression plasmid per plate was added. Cells were incubated for 1 hour at 37°C in a humidified 7% CO2 incubator. Following the 1 hour incubation, 7 ml of DME-8+2 containing 52ug chloroquine per ml was added to the cells, and they were returned to the incubator for five hours. The media was next removed and replaced with fresh DME-8+2. Cells were incubated for 40 hours, then harvested. 36 Figure 8. Construction of PGHSO, expression plasmids. A) pSVT7-PGHSm, is the clone that is currently used in our laboratory as a contI’Ol: B) The cDNA insert was digested from the pSVT7 plasmid with Sall and subcloned 1n“) pSVL. - C) To shorten the 3’UTR of PGHSM, the pSVL-PGHSOV plasmid was digested With DraIII, blunt ended, Sall linkered and religated to form pSVL-PGHSOVA3’UTR. 37 W 19.01.91 A) pSVW-PGHSov PGHSov coding region i I\\\\\\Y\\\\\\\\\V : i pSVT7 53” Dralll 53" B) pSVL-PGHSov PGHSov coding region i 1\\\\\\\\\\\\\\\\N : i pSVL Sall Dralll Sail Accl Accl C) pSVL-PGHSOVA3'UTR PGHSov coding region i m\\\\\\\\\\\\\\‘r : pSVL sit" Sall Accl Accl FIGURE 8 38 To harvest the cells, the media was removed, and the cells were washed with 5 ml of ice-cold phosphate buffered saline (PBS). Cells were removed by scraping in 5 ml of PBS with a rubber policeman and were collected by centrifugation at 2000 rpm for 5 minutes at 4°C. Cells were then gently resuspended in 0.5 ml PBS, transferred to microcentrifuge tubes and centrifuged for 30 sec. PBS was removed from the cells, and the cells were immediately frozen in liquid nitrogen. Comparison of the cyclooxygenase activity of freshly harvested cells and the activity of frozen cells showed that freezing had no effect. Comp_arison of kinetic properties of PGHSmn-l pnd -; For cyclooxygenase and peroxidase assays, microsomes were isolated from the transfected cells (66). To isolate microsomal membranes, frozen cells were thawed and resuspended in 0.1 M Tris, pH 7.5 and sonicated. The sonicated cells were centrifuged at 10,000 rpm for 10 minutes at 4 °C. The supematants were transferred to polyallomer ultracentrifuge tubes and centrifuged at 40,000 rpm for 1 hour. Microsomal pellets were resuspended in 100 mM TrisCl, pH 7.5 (50 pl per plate of cells). Cyclooxygenase activity was determined at 37°C by monitoring oxygen uptake using an oxygen electrode (Yellow Springs Instruments, Model 53). Reactions were initiated by the addition of microsomal enzyme to a mixture containing 3 ml of 0.1 M Tris, pH 8.0, containing 1 mM phenol, 85 pg hemoglobin (as a source of heme) and 100 pM arachidonic acid (74). One unit of cyclooxygenase activity is defined as that which will catalyze the oxygenation of 1 nmol of arachidonate per minute per mg of microsomal protein under standard assay conditions (66). K“1 values were determined using 1.25-100 pM arachidonic acid. 39 Hydroperoxidase assays were performed spectrophotometrically using a Perkin- Elmer model 552A Double Beam UVN IS spectrophotometer. The reaction mixture consisted of 100 mM TrisCl, pH 7.2 containing 5.6 mM guaiacol, 5-100 pg microsomal protein and 1 pM hematin in a total reaction volume of 0.3 ml. Reactions were initiated by the addition of H202 to a final concentration of 400 pM and the product formation was monitored by measuring changes in absorbance at 436 nm (74). Manson of kinetic constants for different fatty acids. Vlam values for cyclooxygenase activity were determined for eicosapentaenoic acid (20:5, EPA), and Km and V“m values were determined for arachidonic acid (20:4, AA) and eicosatrienoic acid (20:3, ETA) for PGHSm-l and PGHSmu-Z. For Vum experiments each fatty acid was used at a concentration of 100 pM. K"1 determinations were performed for ETA at fatty acid concentrations ranging from 125-100 pM. Segpence comparisons. Sequence comparisons performed using the Bestfit sequence alignment programs (75, 76) from the Genetics Computer Group, Inc. sequence analysis software package (77). This software was run on a Vax 8650 at the Michigan State University computer laboratory. 40 Results Northern blot analysis of mouse mRNA, (Figure 9), revealed the presence of two distinct bands that hybridized with the ovine PGH synthase cDNA. One at 2.7 kb corresponded in size to the mouse homologue of the sheep vesicula gland PGH synthase and a second unknown band was present at 4.5 kb. Hybridization to the larger species was enhanced when the stringency of the hybridization was decreased (from 50% formamide to 30% formamide) suggesting that the larger mRNA was related but not identical to the known murine PGH synthase. Based on the results of the Northern blots, a lgth library constructed from serum- stimulated Swiss 3T3 cells was screened using the ovine PGH synthase cDN A as a probe. Initial screening was performed at low stringency (30% formamide) since we were interested in identifying clones corresponding to the 4.5 kb message. Following initial identification of positive plaques at low stringency, the plaques were rescreened at both low and high stringency (30% formamide and 50% formamide, respectively) and a number of clones were identified that hybridized preferentially at low stringency. Restriction analysis of these cDN A clones revealed a common restriction pattern different from that of PGHSm-l cDNA clones (Figure 10). This suggested that these clones were probably derived from a single mRNA species different from the PGHSmu-l. A single phage clone, Agt10-15.2, which appeared to contain the largest cDNA insert was selected for further analysis. When the cDNA from this clone was hybridized to a Northern blot of mouse 3T3 cell mRNA, it hybridized to a single mRNA of 4.5 kb in length, confirming that it was indeed derived from this mRNA. Digestion of this cDNA clone, kgt10-15.2, with EcoRI yielded 5 fragments (1.64, 41 Figure 9. Northern blot analysis of mouse 3T3 cell mRNA hybridized with a 1-7 1‘” EcoRI fragment from sheep cDNA PGH synthase at high (50% formamide) and low (30% formamide) stringencies. 42 High Low FIGURE 9 43 5:259... 853.2 :5 28m 2: ee 8:82 Ba one 2: ”:53... 22% NEE”: Ea $2.25.. as were; 5.338: .3 2:3... «.mIGQ doses. .832 9:03 9. mad «6 u q u q i] E E E 9. me p E 9. F. E 9. m: E meon. mono—2 no.3. 9.63 e. 3. 5. an. a d a l E E E e. 2., E E 44 1.25, 1.10, 0.3 and 0.28 kb, Figure 10). These pieces were subcloned into the vector M13mpl9 for sequencing. Sequencing revealed that the th10-15.2 clone coded for a protein with an open reading frame of 604 amino acids with a 5’ untranslated region of 48 nts (Figure 11). The 3’ untranslated region was about 2.5 kb in length and contained 11 capies of the sequence AUUUA, a sequence which has previously been associated with mRNA that is rapidly degraded (71). One fragment (0.28 kb) contained a string of at least 50 adenosine residues adjacent to the sequence of the EcoRI linker used for the construction of the cDNA library; this indicated that our clone contained the 3’ end of the cDNA. The sequence of kgt10-15.2 was homologous with our PGH synthase cDNA. The proteins coded for by the open reading of lgt10-15.2 was 64% identical with the previously isolated PGH synthase cDNA (66). Using a Dayhoff table for permitted amino acid changes (78) the sequences are 79% similar at the amino acid level. The nucleotide sequences of lgt10-15.2 and the mouse PGH synthase were only 68% homologous. Many amino acid sequences thought to be important in the function of PGH synthase are conserved in this newly isolated clone including: a region with homology to the EGF receptor, histidines thought to be involved in the binding of heme, the aspirin binding site, the putative active site, and the glycosylation sites (Figure 11 and Table 3). The major differences between the two sequences were near the amino and carboxyl termini. The amino terminus of the murine PGH synthase contained a hydrophobic sequence of 17 amino acids (bolded) which is absent in the lgt10-15.2. The new clone, on the other hand contained an 18 amino acid insert near the carboxyl terminus (bolded) that is not present in the murine PGH synthase cloned previously. The carboxyl insert is in a region Figure 11. Comparison of the amino acid sequences of PGHSm-l and PGHSm-Z. Areas of conservation include the axial and distal heme binding sites (underlined), the active site tyrosine (bold), 3 glycosylation sites (bold) and the aspirin acetylation site (bold and asterisk). There are two main regions where the amino acids differ between the two proteins. PGHSmu-Z lacks 17 amino acids at the amino terminus that are part of the signal peptide of PGHSm-l. PGHSm-Z also contains an 18 amino acid insertion near the carboxyl terminus in an area of low sequence similarity between the two proteins. PGHS-l 1 PGHS-Z 1 51 34 101 84 150 134 200 184 250 234 300 284 350 334 400 384 450 434 500 484 550 534 587 584 46 KSRRSLSLWFPLLLLLLLPPTPSVLLADPGVP SPVNPCCYYPCQNQGVCV 5 0 ............... MLFRAV. .LLCAALGLSQAANPCCSNPCQNRGECM 33 RFGLDNYQCDCTRTGYSGPNCTIPEIWTWLRNSLRPSPSETHFLLTHGYW . I: I. I. IIIIIII: I. III. II: I:l I. .I: :III STGFDQYKCDCTRTGFYGENCTTPEFLTRIKLLLKPTPNTVHYILTHFKG LWEFVNA.TEIREVLMRLVLTVRSNLIPSPPTYNSAHDYISWESFSNVSY :|::l|. .|:l.: l:.||| ll ||.|lI||| .:| l||.|||:|| VWNIVNNIPFLRSLTMKYVLTSRSYLIDSPPTYNVHYGYKSWEAFSNLSY YTRILPSVPKDCPTPMGTKGKKQLPDVQLLAQQLLLRREFIPAPQGTNIL ||| ll.|:.l||||l|.ll.|:l|| . : :.:ll|lll||.||l.i:: YTRALPPVADDCPTPMGVKGNKELPDSKEVLEKVLLRREFIPDPQGSNMM FAFFAQHFTHQFFKTSGKMGPGFTKALGHGVDLGHIYGDNLERQYHLRLF lllllllllllllll. |.||l||::l||||||.||||:.|:l|..l||l FAFFAQHFTHQFFKTDHKRGPGFTRGLGHGVDLNHIYGETLDRQHKLRLF KDGKLKYQVLDGEVYPPSVEQASVLMRYPPGVPPERQMAVGQEVFGLLPG IIIIIIIII: :IIIIII. I. I I III :l.: I: IIIIIIIII: II KDGKLKYQVIGGEVYPPTVKDTQVEMIYPPHIPENLQFAVGQEVFGLVPG LMLFSTIWLREHNRVCDLLKEEHPTWDDEQLFQTTRLILIGETIKIVIEE ||::.llllllllllll:||:|||.|:l||||l|.|||||||l||||ll: LMMYATIWLREHNRVCDILKQEHPEWGDEQLFQTSRLILIGETIKIVIED YVQHLSGYFLQLKFDPELLFRAQFQYRNRIAMEFNHLYHWHPLMPNSFQV llllllll :. ||l|||||| Jlll |||| l|| lllllll: I: l YVQHLSGYHFKLKFDPELLFNQQFQYQNRIASEFNTLYHWHPLLPDTFNI GSQEYSYEQFLFNTSMLVDYGVEALVDAFSRQRAGRIGGGRNFDYHVLHV IIII: ill: I. I: |::. I: I I II III: :llll l l EDQEYSFKQFLYNNSILLEHGLTQFVESFTRQIAGRVAGGRNVPIAVQAV AVDVIKESREMRLQPFNEYRKRFGLKPYTSFQELTGEKEMAAELEELYGD l . |.:l|||:.|.:|||||||:lIll|||:l|||||l||||l..l|:l AKASIDQSREMKYQSLNEYRKRFSLKPYTSFEELTGEKEMAAELKALYSD * IDALEFYPGLLLEKCQPNSIFGESMIEMGAPFSLKGLLGNPICSPEYWKP l|.:|:|l:ll:|l..l:.l||l.|:|:||||||l||:|||||l|:|||l IDVMELYPALLVEKPRPDAIFGETMVELGAPFSLKGLMGNPICSPQYWKP STFGGDVGFNLVNTASLKKLVCLNTKTCPYVSFRVPD ............. llllltlll.::|||l:..|:l l-l.||:.|l.|.l STFGGEVGFKIINTASIQSLICNNVKGCPFTSFNVQDPQPTKTATINASA ....YPGDDGSVLV. RRSTEL 602 : IL llllll SHSRLDDINPTVLIKRRSTEL 604 100 83 149 133 199 183 249 233 299 283 349 333 399 383 449 433 499 483 549 533 586 583 FIGURE 11 47 Table 3. Comparison of PGHSm-l and PGHSm-Z PGHS-l PGHS-Z Other Names and Homologues cyclooxygenase, Cef-l47, sheep vesicular TIS-lO, gland PGH miPGHSd, synthase mRNA 2.7 Kb 4.8 Kb Protein 602AA’s 604 AA’s, MW: 69,054 MW=69,093 mRNA Induced by- Serum T TTTT TPA T TTT PDGF T II T SRC N.D. TTT cAMP I III HBGF-l I N.D. IL-l T no effect LPS I T TT Glucocorticoids +- iii Conserved regions EGF homology domain AA’s 34-81 AA’s 17-64 Heme binding His311,209,390 His295,l93,374 Aspirin binding site Ser532 Ser516 Active Site Tyr387 Tyr 371 Glycosylation sites Asn70,106,l46,41 Asn53,130,396,5 2 80 Signal Peptide Yes, 26 AA’s Yes, 17 AA’s 3’-UTR AUUUA sequences No Yes Gene Structure 11 exons 10 exons 22 Kb 8 KB 48 of low similarity between the two sequences. Based on the similarities between the two sequences, we called the newly isolated cDNA clone PGHSmu-Z while the classical murine PGH synthase now became PGHSmu-l. The sequence similarities between PGHSm-l and PGHSm-Z suggested that PGHSmu—Z might have catalytic properties similar to PGHSmu-l. In order to determine if this was the case, the two cDNA’s were subcloned into the expression vector pSVT7. Preliminary studies with PGHSmu-Z were performed by ligating a 2.6 kb EarI/Sall fragment of PGHSm-Z along with a 5’ oligonucleotide linker adapter into pSVT7 (Figure 5B). The recombinant pSVW-PGHSm-Z was used to transiently transfect cos-1 cells Following harvest of the cells, cyclooxygenase and peroxidase assays were performed Cells transfected with pSVW-PGHSmu—Z did not exhibit cyclooxygenase or peroxidase activity above that of sham u'ansfected cells (Figure 12A). Northern and Western blot analysis revealed that very little message or protein was produced by the transfected cells. Cells which were transfected with the ovine PGH synthase as a control, had cyclooxygenase and peroxidase activity comparable to that normally seen in our laboratory (Figure 14A) (74). We hypothesized that the PGHSmu-Z message was undetectable because it was rapidly degraded in the cos-1 cells due to the multiple AUUUA sequences in the 3’-UTR. To eliminate the majority of the 3’-UTR, PGHSm-Z was shortened by digesting with EcoRV, which removed 626 bases of 3’-UTR. We decided, at the same time, to add the 5’-UTR from the ovine PGH synthase to PGHSm-Z (Figure 5C,D). The ovine PGH synthase had high activity, which we hypothesized might be attributable to this 5’-UTR. We called this clone pSVT7-PGHSm-2A3’UTR. When COS-1 cells were transfected with 49 85588 x2 95 x8 58 830:8 8:. 2833 8:. 88:3 «PPM 2: .23 <29o 388 2:8: 2: Basso «5.2.88-8? O .8558.“ x8 93 x8 8:38 as 2833 m5 45% 25 8:233 <28 388 388 as .8 323:8 388-82 a $5588 88588 98 8888838 £8 8338a : .CQSBE So 5 8m: 5583 838.538 SmEOA-E.>mQ 35.8: of. 2 852.8 3.8582. ES omucowmxcotmu 8.89.. new 5558.: c5838 smlwm .3 953.”— 82528 x8 Ea x8 58 8838 25:3 3:. 8.5% 88:8 a 2“ ”ES. 8:5 2.. 859:8 82.27.282-158 6 .3338 X2 8: :5 >558 X00 88298 05a 5583 55. .MHD.m 628:8 a 8a ”2.3% 258 2: 858:8 EDEdTEmIBHqu 8 8383855. 83 A5588 88588 05 :5 538a x00 8838a : 25.3% w=2 a use :5on w5e8 Tasmmmum 2: .Eb.m 058 65 858:8 TaemmOm-E.>mn -2 82.58 88582. 53 088385.98 1889. 6:: 85.8... 868.55 72.935 .3 0.52..— .588 x8 Ea x8 58 8:88 a 8.5m a 22 8: 2e 9; NEE 2:5 2: Bases N-.am:8-§ma 6 355.8 x2 8: x8 .222. 8388 as : MES 85:2... a 92 e5 8.9m 2:8 as 859:8 53.28.588-58 8 538a 90% 38588 .5: AXOUV 3888838 8:50: 8: MHDE 098 a 98 2258 05 5 8&8 8588 NaemmOm 05 68 .m 05 3 0:88 855. 058888820 8 858:8 853 ~-_.am0m-E.>mn 2 823.8 0828:2— EE 088388.93 83.0.. ecu 858... 88835 «gimme.— .N— 28E 50 2.8.5.2.: .3222. :3: 3:2 x0500 2: .o .022.- 2: 2a 3:2 XOLROO n 9555391.: 8.2:: « 052.533.“ 8.2:: p N ad 56 w adv 0 n6 win win m fin ads mfia maxm .02 nxOnROO axon. . x8 mm...H.HH.W....M..H.W,W_...H..m.fl_w..w.man“ 4m.“ 8. WE .0 v :52 .626 .48: $1..m_._..m_..m,_m..mflu. mm . .s. i an 33 8 .. Si 2510 ...>m _¢S.2uzsofifl 2 2.... Bang. n 0." odw mdu n .. 0 ad r N .. o v.5 w mnxm .02 nxgxxoo a x8 . x8 ................................................... m.Phd.nd =mmmumm .‘ ...... :53: “m8 :S.nS.:Emzo:..M >m . :52 =3552 . u... . p: Yucatan. >m 0 ad ad r odu a .. o 0 men .oz .5258 .5: .8 mm_._._>.._.0< :Shfimumawnmmumgm . 01.0130... _ as." "35.88% 2 «.aEmzom.b_.>m whoafimzoo E .5505 mg my—DOE S EOE 51 pSV'I'7-PGHSmu-2A3’UTR, the microsomes again exhibited no cyclooxygenase or peroxidase activity (Figure 12B). Although much of the 3’UTR was removed when making pSVT7-PGHSmu-2A3’UTR, there were still 7 AUUUA repeats in what remained of the 3’UTR of PGHSmu-Z. We decided to eliminate the remainder of these sequences. We digested a second time with Accl, which cuts 11 bases past the translational stop codon. We also decided to simultaneously try a second expression vector, pSVL, which is similar to pSVT7 except that it utilizes a different promoter, the SV40 late promoter instead of the SV40 early promoter. The new expression vector we had constructed was called pSVL-PGHSm-Z (Figure 5E). Transfection of cos-1 cells with this vector finally resulted in expression of the PGHSmu-Z protein. We found this protein to possess both cyclooxygenase and peroxidase activities (Figures 12C and 15A). Interestingly, when the truncated PGHSmu-Z was subcloned into pSVT7 and used to transfect cos-l cells, there was no measurable cyclooxygenase or peroxidase activity (data not shown). We have no explanation for this, but it does point out that one must determine empirically the best vector for expression of each cDN A. The clone pSVL-PGHSmu-Z was used for all further studies on PGHSmu-Z activity. To directly compare the properties of PGHSm-l and PGHSmu-Z, analogous plasmids were constructed containing PGHSm-l (Figure 7). The first PGHSmu-l expression plasmid contained the S’UTR from sheep PGH synthase followed by the PGHSm-l coding region (Figure 7C). The PGHSmu-l cDNA fragment used in this construction ended at a Xbal site 643 bases past the translational stop codon. The clone constructed in this manner was called pSVTV-PGHSmu-l. Assay of microsomes harvested from cells transfected with this pSVT7-PGHSmu-l revealed that this clone expresses PGH 52 Figure 15. Oxygen electrode traces of cyclooxygenase activity of microsomes from COS-1 cells transfected with the PGH synthase expression plasmids. The COX activity is measured as the initial slope following addition of the enzyme to the reaction mixture. Enzyme was added at the point indicated by the arrow; the size of the small peak following enzyme addition directly relates to the time for enzyme activation. There is a very small peak for PGHSmu-Z and a larger peak for PGHSm-l. A) pSVL-PGHSmu-Z B) pSVT7-PGI-lSmu-1A3’UTR 53 PG HSmu-1 A. PGHSmu-2 time—b .3963 cozommillv .i .i#v.l.lt Tile 1". Illicl‘y . 0. . .. t .. . o a . o - FIGURE 15 54 synthase activity efficiently (Figure 13A). However, the specific cyclooxygenase activity of PGHSm-l was about 1/3 that found for the ovine PGH synthase (Figure 14). The peroxidase activity of this clone was too low to be measurable (Figure 14A). To obtain better expression of PGHSm-l we decided to remove a portion of the 3’UTR and subclone the coding region into pSVL as we had the PGHSmu-Z (Figure 713). Although the 3’UTR of the PGHSmu-l cDNA does not contain any copies of the AUUUA sequence (66), we did not know what other factors might be contained within the 3’UTR which would regulate stability of the PGHSmu-l message. The resulting clone was called pSVL-PGHSmu-1A3’UTR. pSVL-PGHSmu-1A3’UTR was used to transfect COS-1 cells and the microsomes assayed for cyclooxygenase and peroxidase activities. As before, PGHSm-l had cyclooxygenase activity but no detectable peroxidase activity (Figure 13B). Because it was easier to amplify pSVT7 clones than pSVL clones, we also subcloned the shortened PGHSm-l into pSVT7, creating pSVT7—PGHSmu-1A3’UTR (Figure 7B). When transfections were performed using this clone, we found that the cyclooxygenase activity was somewhat higher than had been previously observed, and that with the increase in expression the peroxidase activity was now measurable (Figures 13C and 15B). The clone pSVW-PGHSmu-l was used for transfection in all further experiments on PGHSmu-l activities. An analogous set of clones was constructed for the ovine PGH synthase which, because of its high activity, was a useful positive control for transfection experiments (Figure 8). The ovine PGH synthase was first removed from pSVT7 (Figure 8A) and subcloned into pSVL, creating pSVL-PGst (Figure 8B). Next, the 3’UTR was removed from the ovine cDNA in pSVL to create pSVL-PGHSOA3’UTR (Figure 8C). We found 55 that both of the pSVL clones had a somewhat lower activity than the original ovine clone, but the activities were not different between the two pSVL clones (Figure 14). Again, it became obvious that the optimal plasmid for expression of each cDNA could not be predicted. Thus we returned to using the original pSVT7-PGHSOV as a control plasmid for all our experiments. Once we had determined that PGHSmu-Z possessed the cyclooxygenase and peroxidase activities associated with PGHSmu-l, we began to examine the kinetic properties of the enzymes. We first compared the Km and Vm, values for the PGHSOV, PGHSmu-l and PGHSmu-Z for arachidonic acid. Vmu values were determined using lOOpM arachidonic acid (Figures 12-14). The Vm, values for the three enzymes expressed in cos-1 cells with arachidonic acid are quite different from each other. The ovine enzyme always exhibits the highest cyclooxygenase activity V,mm (92.5 nmol arachidonic acid (AA) consumed/min/mg). The value for PGHSmu-l and PGHSmu-2 in transfected cells is roughly 1/3 of that seen for the ovine PGH synthase-1 cDNA (29.5 and 23.0 nmol AA/min/mg, respectively). Whether this is due to differences in efficiency of expression of the three enzymes, or is due to actual differences in kinetic efficiencies of the two enzymes is not known. K,11 values were determined by measuring activity with arachidonic acid concentrations from 1.25 to 100 pM. We found that there were no significant differences in the Km value for any of the PGH synthase enzymes examined. The KIn for PGHSov-l was 6-8 pM (66); for PGHSmu-l, l.1-3p_M (Figure 16A); and for PGHSmu—Z, 2.5-4.5 pM (Figure 16B). The maximum peroxidase activity for PGst (29.9 nmol AA/min/mg) was higher than that seen for either of the murine clones (Figures 12-14). The peroxidase activity 56 N-_._.Hmmom a .coambE 295m .3 sauce v.83 3:5 43825;" 0 T mIOm 2 . . . . 5 .3 32865 m_ 68.35 05 .8 EM 0 . Eu 3 $8 uxcsmcoéoaocfi 2: .3 @983 mm 8mm .23 3.82:an .5. ~52ng ecu 16:0.” .5.— ..v— 2: he gag—EMSG 4m: ”Mo—“mi 57 0.. wan-0E 23: 73:: m6 m... to me «.o to o.o to- No- 0.9 «.0. ad. a... no to Go No .6 ad to- «a- n6- «6. n I n I by I PL b F b I L IL I h I h I h P I .— I b I I I h I n L I h r b I - rh 18.0 I .35 1 0 .mod .2: mNnEx .2: o 9:45. NbEmIGn. 72.5101 It I e m O .86 n O .35 X . W W 15... X .36 n. V . m. a I .8... . nu .86 .A A o o . .86 M o .soo 0 road rhod 58 of PGHSmu-Z (13.8 nmol/min/mg) was somewhat higher than that of PGHSmu-l (10.6 nmol/min/mg). The cyclooxygenase/peroxidase ratio for PGHSm-Z (2.3) was lower than that of PGHSOV (3.6) or PGHSmu-l (3.3), suggesting that the PGHSm-Z may have relatively more peroxidase activity than PGHSmu-l. We next compared the ability of PGHSmu-l and PGHSm-Z to utilize different fatty acids as substrates for cyclooxygenase activity. Two different fatty acids were examined: eicosatrienoic acid (ETA) and eicosapentaenoic acid (EPA) (Figure 17). ETA is a common mammalian fatty acid (79), the precursor of prostaglandins of the l-series (e. g. PGE,, PGDI). EPA is a fatty acid found in fish oil that has been implicated in the reduction of cardiovascular disease (80). As shown in Table 4 and Figure 18, ETA is about half as efficient for PGHSmu-l as is arachidonic acid. The ratio of the Va"m values for 100 11M ETA and 100 11M AA for PGHSmu-l is 0.5. PGHSmu-Z, however, utilizes ETA more efficiently than arachidonic acid; the ratio of the Van, values for 100 pM ETA and 100 ul_vl_ AA for PGHSmu-Z is 1.5. The K“I values for PGHSmu-l and PGHSm-Z with ETA (5.6 11M and 6.2 pM respectively) (Figure 19) were not significantly different from each other or from the K"1 of these two enzymes with arachidonic acid (3.0 pMPGHSm-l; and 2.5 pM, PGHSm-Z) (compare Figures 16 and 19). EPA is a very poor substrate for PGHSmu-l (Figure 20 and Table 4); cyclooxygenase activity with this substrate was barely detectable. When we could detect PGHSmu-l cyclooxygenase activity with EPA, the Vmu ratio between 100 pM EPA and 100 pM AA was 0.04. EPA, however, is readily utilized by PGHSm-Z; the V,1m ratio between 100 FM EPA and 100 ply! AA ranged from 0.34-0.37. 59 Figure 17. Structures of fatty acid substrates tested for cyclooxygenase activity with PGHSm-l and PGHSm-Z. ARACHIDONIC ACID (20:4, AA) COOH EICOSATRIENOIC ACID (20:3, ETA) EICOSAPENTAENOIC ACID (20:5, EPA) COOH FIGURE 17 61 Vmax (nmol AA/min/mg) V V max = 0.53 max = 1.51 V V max max Figure 18. Comparison of the Vm, for PGHSm-l and PGHSm-Z using arachidonic acid (AA) or eicosatrienoic acid (ETA) as the substrate. 62 Table 4A. V“ml ratios for PGHSm-l and PGHSm-Z using different substrates PGHSm-l PGHSm-Z Table 4B. Vm,‘ values for PGHSm-l and PGHSm-Z for individual experiments PGHSm-l 35.0 18.5 23.6 13 1 78.7 2 8 54.1 0.0 PGHsm-z 29.1 44.0 12.4 19.4 47.9 16.3 32.5 12.1 63 $2.32 a 72%? 2 62$“:on 295m .3 5596 203 8:5 4382:: 2: E 332?: mm 5M 2: 823 e052: 3593503054 65 .3 @283 a San— .QFm: Eon 9652.883 .5. Nuammvm ecu fiasmmwm 3.. av. 2: he :ozafiEcfloa .3 Earn aw NGDGE 25:: 25:: mo vo no mo Po oo .o- «o- 9° co no mo .o oo .9 N? r p . u p u p b . p P p p . . F» p > p . n p b L t n p L w o . 1 .2: QmuEX .2: m 35. #3on“. . msEmIOd I. I. wvod nlu I o m . x x v v 186 m w M M. . .M M .85 [°F.° I N°.° 65 [3 EPA Vmax (nmol AA/min/mg) 1 llJLilllllJllllllllllllllllllllLllllllll] v EPA v EPA _"EX____ = 0.04 L = 0,34 max max Figure 20. Comparison of the V...“ for PGHSm-l and PGHSm-Z using arachidonic acid (AA) or eicosapentaenoic acid (EPA) as the substrate. 66 Discussion Despite our early work with Southern blot analysis which indicated that only a single PGH synthase gene existed, Northern blot analysis led us to suspect that a second 4.5 kb mRNA might be related to the murine PGH synthase. A cDN A derived from this mRNA was isolated using differential screening techniques. Sequencing of the clone confirmed that it was closely related to the murine PGH synthase. During our initial characterization of this cDNA, this second gene was discovered and reported by two other groups (26, 64). The chicken form of PGH synthase-2 was isolated by Xie et al. (64) from a cDNA library made from chicken embryo fibroblasts transformed by a temperature sensitive Rous sarcoma virus. This cDNA coded for a protein of 603 amino acids which exhibited 61.3% identity with the ovine PGH synthase (Table 5). The mRNA for this chicken gene was unusual in that it existed in two different forms. At non-permissive temperatures, the predominant form of the clone contained an unspliced intron near its 5’-end. At permissive temperatures, active srcmo protein is expressed and cells assumed a transformed phenotype leading to removal of the intron and translation of the PGH synthase—2 mRNA. As we found with our mouse PGH synthase-2 cDNA, the chicken cDNA lacked hydrophobic residues near the amino terminus of the predicted protein, known to comprise the PGH synthase-1 signal peptide. The chicken cDNA also contained an additional 18 amino acids near the carboxyl terminus not found in PGH synthase-1. The nucleotide sequence of the chicken PGH synthase-2 cDNA contained sixteen AUUUA sequences in its 3’UTR. A second clone for the murine PGHSm-Z was isolated by Kujubu et a1 (26). These authors initially identified a clone which they called TISlO, as one of a group of 67 a...» .. ed» .. Q; 3.x .- v.3 age .. 3% fine as .. h? 3:. fit. .. ad. «.3 ed. anew -. «Sb 3:. v.2. wda -. «.8 «.3 3» new as 3:. 3:. 35 a 35 «5:930 «9252 Nee—5.: ~83: 33.—m 23:5: brazfimoe 555388 .2325 5253* beseeas .8252 355388 regime .225: $2818 regime 43.52 gauge regime .2025 b25388 3:22:54. .225: .3225: .3 ES; 250 «ma—.3? =0.— 2: .8 23:32 6565 95,—. .m «35. 68 immediate early genes induced by the addition of phorbol esters (TPA; hence IPA _I_nducible Sequence) to density-arrested Swiss 3T3 cells. After sequencing they found that T1810 coded for a protein with an open reading frame of 604 amino acids with high sequence similarity to the ovine and murine PGH synthases. The authors further isolated genomic clones and determined that the transcriptional start site was located 44 nucleotides upstream of the translational start site. Additional experiments by these authors showed that PGH synthase-2 (T1810) was highly induced by EGF, forskolin and serum, as well as TPA. They found that the second PGH synthase was produced in a tissue specific manner, and was expressed in only a few of the many cell lines they tested. More recent work in our laboratory suggests that the second PGH synthase is expressed only after stimulation by cell specific factors and thus may be more ubiquitous than originally thought (Kraemer and DeWitt, unpublished results). The results of these early studies suggested that the newly identified PGH synthase is "inducible" while PGH synthase-1 seems to be constitutively expressed. Expression of PGHSm-Z in cos-1 cells by us and others (94) revealed that it had both the cyclooxygenase and peroxidase activities associated with PGHSmu-l and probably functions as a PGH synthase enzyme (Figures 12-14). It was not without great difficulty that we were able to express PGHSmu-Z in vitro. These difficulties were likely due to the instability of the PGHSmu-Z mRNA; deletion of the entire 3’ untranslated region including the multiple AUUUA sequences finally allowed us to efficiently express this enzyme (Figure 5E). The instability of the PGHSmu-Z mRNA suggested that this gene is expressed only transiently, and this hypothesis has been confirmed in 3T3 cells (see Chapter 4). We had little difficulty expressing the murine PGH synthase-1. This cDNA 69 lacks the deStabilizing AUUUA sequences in its 3’UTR and appears to produce a much more stable message. Removal of the 3’UTR did not alter the efficiency of expression of either the ovine or murine PGH synthase-1 (Figures 13 and 14). However, we found empirically that the two genes were expressed more efficiently from different promoters. PGHSm-Z was best expressed in the pS VL vector which contains the SV40 late promoter, while the ovine and murine PGH synthases-1 were better expressed in pSVT7 which contains the SV40 early promoter. There is little difference in the affinity of any of the PGH synthases for arachidonic acid; the Km values for the cyclooxygenase activities of PGHSov-l, PGHSmu-l and PGHSm-Z are not significantly different from one another (68 11M, 1-3 11M, and 2.5- 4.5 11M respectively) (Figure 16). There is, however, a difference in the vmax for the murine enzymes compared to the sheep enzyme (Figures 12-14). The Va,“ for the sheep enzyme is two to three times as great for the sheep enzyme as for either of the two mouse enzymes, while each of the mouse enzymes are expressed to about the same level. Determination of whether this difference is due to a more efficient expression of the sheep enzyme, or due a catalytically more active sheep enzyme will require purification of these enzymes. While PGHSov—l had greater peroxidase activity than PGHSm-l, the cyclooxygenase/peroxidase ratios were similar for the PGHSo,-1 and PGHSmu-l clones and lower for PGHSmu-Z. This suggests that PGHSmu-Z may have more peroxidase activity than PGHSm-l. This difference may, however, be due to an artifact of the expression system. In our system, the greater peroxidase activity of PGHSmu-2 appears to lead to more rapid activation (Figure 15). Whether the small difference in the peroxidase 70 activities of the two isozymes would be of physiological significance is unknown. The two murine PGH synthase enzymes have quite different abilities to utilize fatty acids other than arachidonate as substrates. Although the KIn values of PGHSmu-l and PGHSm-Z for ETA are very similar, 5.6 11M and 6.2 11M respectively, (Figure 19), there is a large difference in their Vm values (Figure 18 and Table 4). Direct comparisons of the same enzyme preparations with the two fatty acids showed that PGHSmu-Z is able to catalyze oxygenation of ETA much more rapidly than is PGHSm-l, suggesting that this enzyme may be more efficient at synthesis of prostaglandins of the l-series. Similarly, PGHSmu-Z will use EPA as a substrate while PGHSmu-l will not (Figure 20 and Table 4). These observations raise interesting questions about the potential function of PGHSmu-Z in viva. As mentioned above, PGHSmu-2 is thought to be the "inducible" PGH synthase. If this is true, and if PGHSmu-Z is induced in ovulation or inflammation, the relative non-specificity of PGHSm-Z might have some interesting implications. Changes in dietary fatty acids, for instance consumption of large amounts of fish oil (EPA) might affect inflammatory responses more dramatically than events mediated by production of prostaglandins by PGHSmu-l. PGHSmu-l is able to utilize primarily one substrate, arachidonic acid. This restricts the types of prostaglandins that PGHSm-l can produce and suggests that cells such as platelets, which express primarily this isozyme (81), cannot utilize EPA or ETA efficiently. CHAPTER 3 PHARMACOLOGICAL CHARACTERIZATION OF PGH SYNTHASE-l AND PGH SYNTHASE-Z Introduction Non-steroidal anti-inflammatory drugs (NSAIDS), used for everything from relief of minor pain to relief from rheumatoid arthritis and prevention of heart attack, are among the most widely consumed drugs in the world (82). NSAIDS have in common one feature, they inhibit prostaglandin synthesis by the enzyme PGH synthase (83). The aim of the experiments presented in this chapter is to compare the sensitivities of PGHSm-l and PGI-ISm-Z to common NSAIDS. It appears that PGH synthase-1 and PGH synthase-2 play different physiological roles. For instance, there is evidence that PGH synthase-2 is only expressed in stimulated cells, such as immune-activated macrophages (53), while PGH synthase-l is constitutively expressed in most tissues (84), including platelets (81). Although the exact roles of these two isozymes is not known, it is likely these roles are different. It is therefore likely that specific inhibitors of the isozymes will be therapeutically useful. Aspirin, the most commonly used NSAID, inhibits PGH synthase by two distinct mechanisms; instantaneous reversible competitive inhibition of substrate binding, and time-dependent irreversible inhibition due to acetylation of a serine residue in the substrate binding region of the enzyme (85). The exact mechanisms of inhibition by NSAIDS other than aspirin are less well understood. Inhibition studies, beginning with 71 72 those presented in this chapter, with both the PGHSmu-l and PGHSmu-Z enzymes may eventually help clarify the structure-function relationships between NSAIDS and PGH synthase inhibition. 73 Methods Expression of PGHSmu-l and -2 in a transient expression system. PGHSmu-l and PGHSm-Z were transiently expressed in cos-1 cells as previously described (Chapter 2). The plasmids used for expression of PGHSmu-l and PGHSm—Z proteins were: pSVT7-PGHSm-1A3’UTR (Figure 7E) and pSVL-PGHSmu-Z (Figure 5E). The plasmid pSVT7-PGHSO, was used as a control for transfection efficiency in each experiment. Transfections were performed by the DEAE dextran/chloroquine method (66). Cells were harvested 40 hours following transfection and frozen in liquid nitrogen for later use. Microsomes were isolated from the frozen cells as needed (66) and resuspended in 0.1 M TrisCl, pH 7.5, at a concentration of 50 pl buffer for microsomes from each plate of cells. Determination of cyclooxygenase activitv. Cyclooxygenase activity was determined by monitoring oxygen uptake at 37 °C with an oxygen electrode (Yellow Springs Instruments Model 53). Reactions were initiated by the addition of microsomal enzyme to a mixture of 3 ml of 0.1 M TrisCl, pH 8.0, containing 1 mM phenol, 85 pg hemoglobin and 100 11M arachidonic acid (74). For IDso determinations 10 11M arachidonic acid was used as the substrate. One unit of cyclooxygenase activity is defined as that amount of enzyme which will catalyze the oxygenation of 1 nmol of arachidonate per minute per mg of microsomal protein under standard assay conditions (66). Inhibition of cyclooxygenase activitv bv non-steroLdgLanti-inflagrmatory dflgg, Indomethacin, acetaminophen, meclofenamate, and aspirin were purchased from Sigma. The s-isomer of ibuprofen was purchased from Aldrich. Piroxicam was a gift of Dr. 74 Thomas Carty at Pfizer. Sulindac sulfide was from Merck, Sharpe and Dome and 6-methoxy-2-naphthyl acetic acid (6-MN A) from Smith Kline Beacham. Flurbiprofen was resuspended in 0.1 M TrisCl buffer containing 1 m_M_ phenol (Tris-phenol), piroxicam was resuspended in acetone; all other NSAIDS were resuspended in ethanol. Inhibitors were added to reaction mixtures just prior to the addition of enzyme. Dose-response curves were performed for flurbiprofen, ibuprofen, meclofenamate, piroxicam, sulindac sulfide and indomethacin (structures shown in Figure 3, families of inhibitors shown in Table 2). Vehicle controls were performed for each inhibition. Mon of cvcloorg/ggrpse activitv mm The time—dependent inhibition of PGH synthase by aspirin (Figure 3) was studied by incubation of microsomal membranes with 200 1114 aspirin at 37°C for varying times. Aliquots of the enzyme were assayed for cyclooxygenase activity at O, 10, 15, 20 and 30 minutes. As a control, each microsomal preparation for each enzyme was incubated in the absence of aspirin at 37°C. Product characterization by thin layer chromatography. Thin layer chromatography was used to examine the effects of NSAIDS on product formation for PGHSm-l and PGHSm-Z. Product formation was tested in the presence of flurbiprofen, indomethacin, and aspirin. Inhibitors were added at a concentration of 100 p_M_ to transfected cells 40 minutes prior to harvest. Cells were then gently harvested by scraping in PBS, and collected by centrifugation at 1200 rpm for 5 minutes. Cells were next resuspended in 0.25 ml DME per plate containing 25 pM [1—"C]-arachidonic acid (New England Nuclear, specific activity 53.0 mCi/mmol) and incubated at 37°C for 15 minutes. Cells were removed by centrifugation at 1200 rpm for 5 minutes, and the supematants were transferred to clean tubes. To each supernatant was added 1.5 ml of ice-cold acetone; 75 after vortexing, the denatured proteins were removed by centrifugation. The supematants were again transferred to clean tubes, and 250 pl of 0.2 N HCl and 2.5 ml of CHCl3 were added. Following mixing, the aqueous and organic layers were separated by centrifugation, and the organic layer was removed and dried under nitrogen. The dried samples were redissolved in CHCI3 and spotted on Silica Gel 60 thin layer chromatography plates (EM Science). The plates were developed twice in benzene: dioxane: acetic acid: formic acid (82:14:1z1) (86). Standards were visualized with iodine vapor. Thin layer plates were exposed to XAR-S x-ray film overnight to allow visualization of arachidonate-derived products formed by PGHSm-l and PGHSm-Z. Prostaglandins were identified by comparison of Rf values with authentic standards. Prostaglandins synthesis was quantitated by densitometry of x-ray films on a Bioimage Visage 110 image analyzer. For determination of the time course of the effects of aspirin, plates of cells transfected with plasmids for expressing PGHSmu-l or PGHSmu-Z were treated with 100 11M aspirin for 0, 10, 20, 30 or 40 minutes 40 hours after transfection. Cells were harvested and product formation was quantitated following thin layer chromatography and autoradiography by densitometry. Dose-response curves for aspirin inhibition of prostaglandin synthesis were conducted essentially as for time course experiments. Aspirin, at concentrations between 0.1 and 500 pM was added 40 hours after transfection. Following a 30 or 40 minute incubation at 37°C, the remaining PGH synthase activity was determined by thin layer chromatography, autoradiography and densitometry. 76 CharacteriLation of the unique prod_pct formed by PGHSm-Z following aspirin treatment. Preliminary experiments showed that PGHSm-Z produced a unique product upon treatment with aspirin. Initially, various monohydroxy arachidonic acid derivatives including 12-hydroxyeicosaten'aenoic acid (12-HETE) and lS-hydroxyeicosatetraenoic acid (15-HETE), were run on the thin layer plates along with the products produced from arachidonate by aspirin-treated PGHSm-Z. The major product of this reaction comigrated with 15-HETE. To determine whether this product was indeed 15-HETE, we isolated this product from thin layer chromatography plates and determined its structure by gas chromatography/mass specrroscopy (gc/ms). Cells transfected with PGHSmu-Z were treated with aspirin for 40 minutes to maximize formation of the unknown derivative of arachidonic acid and the cells were incubated with unlabeled arachidonic acid. Thin layer chromatography was performed as outlined above. Additional plates of transfected cells were incubated with [l-“C]-arachidonic acid and the products from these reactions applied in separate lanes to allow the localization of the unknown compound. The silica gel from this area was scraped from the thin layer chromatography plate into a silanized glass tube. An authentic 15-hydroxy-eicosatetraenoic acid (IS-HETE) (Cayman) control was also prepared. This control was extracted and derivatized in parallel with the unknown sample and the sample prepared without arachidonic acid using the same reagents. The silica gel was extracted twice with 2 ml of methanol, which was then pooled and dried under nitrogen. The residues were resuspended in ethyl acetate and extracted twice with 1 ml water. The ethyl acetate was itself then dried under nitrogen. The extracted products and control sample were treated with ethereal diazomethane to methylate any carboxylic acid groups present. The ether was removed, and the sample 77 was next treated with bis(trimethysilyl)trifluoroacetamide containing 1% trimethylchlorosilane for several hours to convert any hydroxyl groups to trimethylsilyl- ethers. The derivatized samples were then analyzed on a Joel JMS AX505H magnetic sector gas chromatograph-mass spectrometer using a DBS-MS, 30 meter capillary column (.32 mm id; .25 pm coating) at the Michigan State University Mass Spectroscopy Center. The chromatography profile was from ISO-325°C at 5°C/min. Electron-ionization mode was employed with an ionization current of 100 mA. 78 Results Effects of NSAIDS on PGHSm..-1 and PGHS “-2 cyclooxygenase activig. A variety of NSAIDS were tested for their abilities to inhibit the cyclooxygenase activity of PGHSm-l and PGHSmu-2. Figure 3 shows the structures of the compounds used and Table 2 lists the NSAIDS and the families to which they belong. We used representative members of the different families of NSAIDS. We observed three patterns of inhibition for the murine PGH synthase isozymes. The first pattern was characterized by drugs that are nearly equipotent inhibitors of PGH synthase-1 and PGH synthase-2. Included in this group are flurbiprofen, ibuprofen, and meclofenamate. The respective [1350 values of these drugs for PGHSmu-l and PGHSm-Z using 10 11M arachidonate are: for flurbiprofen, 0.5 vs. 2.0 11M (Figure 21, Table 6); for ibuprofen, 14 vs. 7.2 11M (Figure 22 and Table 6); and for meclofenamate, 2.0 vs. 13 p_M_ (Figure 23, Table 6). Two out of three of these drugs are better inhibitors of PGHSm-l, but the differences in potencies (2-4 fold) are probably not significant. Another compound that we tested which has not been used therapeutically, but nevertheless which belongs in this first category is docosahexaenoic acid (22:6). This n-3 fatty acid competitively inhibits cyclooxygenase activity presumably because it is non-metabolizable, but can still bind the PGH synthase active site. As shown in Figure 24 and Table 6, the IDws for docosahexaenoic acid are about 10 11M for both PGHSm-l and PGHSm-Z. A second group of NSAIDS are those that inhibit the PGHSmu-l much more effectively than PGHSm-2. These include indomethacin, sulindac sulfide, and piroxicam. The respective IDSOS for these drugs for PGHSm—l and PGHSm-Z using 10 11M 79 100‘. i 80- a?" .1 IE - ‘6 .. < 60— -2 X - mu 8 _ = 2.11114 :‘é’ 40‘ PGHS .E _, - - no °\° _ 20- 0 I ITIITIII I I IIIIIII I IIIIIIII T I ITIIII] - + PGHS -2 3:: [Ill .5 4 + to -160 M 2 - + + ‘1‘ 50 - u 60— X X - 8 - i To _ I + :- 4 “ ' E 0 - PGHSmu 1 20— 0 rfittrfltl T 1111111] I 1111111] I IIIIIIII 0 10'8 10'7 10'6 10'5 10'4 ‘93 [Indomethacin], M Figure 25. Inhibition of PGHS,“ cyclooxygenase activity by indomethacin. 100% activity is the activity of the PGHSmu isozyme in the presence of 10 11M arachidonic acid and vehicle. 100% activity for PGHSm-l is 23.0 nmol AA/min/mg. 100% activity for PGHSm-Z is 10.9 nmol AA/min/mg. 86 ID50 = 14 PM "/0 Initial COX Activity O 0 ‘I I I IIITIT I I I IIIIIrfi I I IIWII I I I TTI—I I I II 1111' 0 10'8 10'7 10’6 10'5 10'4 [Sulindac sulfide], M. Figure 26. Inhibition of PGHSm cyclooxygenase activity by sulindac sulfide. 100% activity is the activity of the PGHSm isozyme in the presence of 10 pM arachidonic acid and vehicle. 100% activity for PGHSm-l is 14.5 nmol AA/min/mg. 100% activity for PGHSm-Z is 17.9 nmol AA/min/mg. >..>:O< X00 .mEC. o\o 87 120 -— : O 100 4 g I >. ‘4/ IE 80 ~ ‘F 2 : PGl-S -2 x _ ID = 2409M G __ 50 0 60 _ E! j 1 g 40 o - PGHS -1 o\ i 20 _: ID50 — 8.7uM 0 I I IIIIII] I I IIIITIT I I IIIIIT] I I IIIIIII I I IIIIIT] 0 10'7 10'6 10’5 10'4 10-3 [Piroxicam], M Figure 27. Inhibition of PGHSm cyclooxygenase activity by piroxicam. 100% activity is the activity of the PGHSmu isozyme in the presence of 10 pM arachidonic acid and vehicle. 100% activity for PGHSm-l is 47.3 nmol AA/min/mg. 100% activity for PGHSm-Z is 27.1 nmol AA/min/mg. °/o Initial COX Activity is 1 Vs] 88 PGHSm-I Est. ID = 2009M 2. / % Initial COX Activity 0) O I .. O _ PGHSmU 2 20 _ I050 = 15 PM o I TTIIIIIr I IIIIIIII I I IIIIIII I ITIIIIII I IIIIIIII o 10'7 10'6 10'5 10'4 10'3 I5'MNA]. M Figure 28. Inhibition of PGHSm cyclooxygenase activity by 6-MNA. 100% activity is the activity of the PGHSM, isozyme in the presence of 10 11M arachidonic acid and vehicle. 100% activity for PGHSm-l is 15.3 nmol AA/min/mg. 100% activity for PGHSm-Z is 22.0 nmol AA/min/mg. 89 100 . PGHS-Z 90" E .2. 8°: > C - — o 70- I- o o . < E 50- . c = . PGHS-1 Z a. _] g 2 50 ‘ . E E 40- ur o - “i at 30- 20‘ 10‘ o I I I T ' I 0 1o 20 30 Time (min) Figure 29. Time course for inhibition of PGHS,,m cyclooxygenase activities by aspirin. Microsomal membranes were incubated at 37°C with or without 200 11M aspirin. At the indicated times, aliquots were removed for cyclooxygenase assays. Remaining activity is given as a percentage of the zero aspirin control at each time. pi SL’ CC f0 UN 90 9O formation of a non-prostaglandin product from arachidonate that was not formed by aspirin-treated PGHSm-l or by PGHSm-l or PGHSm-Z treated with other NSAIDS. The novel product was less polar than prostaglandins (Figure 30), but more polar than arachidonate. Thin layer chromatography was used to characterize the unique non-prostaglandin product formed by aspirin-treated PGHSmu-Z. Comparison with the monohydroxy standards 12- and 15-hydmxy-eicosatetraenoic acid (HETE) established that the product co—migrated with 15-HETE in our solvent system. To further characterize the product, gas chromatography/mass spectroscopy (gc/ms) was performed on the material co- migrating with lS-HETE. A control sample was prepared by incubating PGHSm-Z transfected cells without arachidonic acid following aspirin treatment. Multiple chromatographic peaks were obtained from cells incubated with and without added arachidonate and extracted from the silica gel plate region corresponding to the unknown compound. However, only a single peak was unique to the sample prepared from cells incubated with arachidonate. This peak co-chromatographed and displayed the identical mass spectrum as the trimethylsilyl-ester trimethysilyl-ether derivative of authentic 15-HETE. The methyl-ester TMS-ether derivative of authentic lS-HETE was not obtained, nor did we observe a chromatographic peak with the spectra of the methyl-ester TMS-ether of 15-HETE (87) in our sample derived from the aspirin-treated PGHSm-Z microsomes, suggesting that our diazomethane reagent had not caused methyl ester formation. However, the trimethylsilyl ether/ester derivative gave an easily interpretable, unequivocal spectrum (Figure 31) with a base ion of (m/z) 225 and ions of (m/z) 374 (M- 90, loss of MgSiOH), 449 (M-lS, loss of methyl), and 464 (M). We cannot determine 91 Figure 30. Product formation by PGHSm-l and PGHSm-Z treated with various NSAIDS. Aspirin, indomethacin and flurbiprofen were incubated for 40 minutes prior to harvest with cells which had been transfected with PGHSmu-l and PGHSm-Z. The ability of the PGH synthase isozymes to synthesize prostaglandins following inhibition by these NSAIDS was determined following the addition of [l-”C]-arachidonic acid (25 p_l\_4, 53.0 mCi/mmol). After incubation with arachidonic acid, the products were extracted and separated by thin layer chromatography (TLC). TLC plates were exposed to x-ray film overnight. nf‘LICMII_O on NEDGI mmm 002. o._ (sz TImGQ 02.22... F m.95... (Zn—E «.mzwn $23.2..— FIGURE 34 120 cDNA (Figure 10), as well as a plasmid containing 2.3 Bsle/Xbal fragment from the murine PGH synthase—1 were used as probes for the run-off assay. Multiple contiguous PGH synthase-2 cDNA fragments allowed us to discriminate between regulation by transcription initiation and/or transcriptional elongation. A slightly larger PGH synthase-1 probe, corresponding to nearly the entire cDN A, was used because smaller probes gave insufficient signal. Figures 35A and 35E show results from nuclear run-off assays of PGH synthase-1 mRN A synthesis. Hybridization to the PGH synthase-l cDNA probe increased 3-fold 3 hrs following serum stimulation. Thus, there was a 3-fold increase in transcription of the PGH synthase-1 gene at 3 hrs; this correlates well with the 3-fold increase in PGH synthase-1 mRNA levels seen 3-6 hrs post serum addition. These results suggest that serum-stimulated increases in PGH synthase-1 mRNA result from increased transcription of the gene. Changes in transcription of the PGH synthase-2 gene were much more dramatic than those observed for the PGH synthase-1 gene (Figures 35A,E). Hybridization to each of the three PGH synthase-2 probes increased within 30 min of serum stimulation to 10- 25 times their initial values (Figure 35C). After reaching its peak at 30 min, PGH synthase-2 transcription gradually returned to near zero time levels by 4 hrs post serum addition. Changes in PGH synthase-2 transcription correlate well with the observed increases in mRNA (Figure 34B,F). Serum stimulates a rapid early increase in transcription of the PGH synthase-2 gene by 30 min, and PGH synthase-2 mRNA levels are maximal only 30 min later. As rates of transcription return to basal levels at 4 hrs, so do PGH synthase-2 mRNA levels. The rapid decline in PGH synthase-2 mRNA levels 121 Figure 35. Transcription of the PGH synthase-1 and -2 genes as determined by nuclear run-off assays. Nuclei from serum-stimulated or serum plus dexamethasone- stimulated 3T3 cells isolated during preparation of RNA for Northem blot analysis (Figure 34) were incubated with 32P-U'I'P for run-off analysis as described in Methods. Labeled transcripts obtained from these incubations were hybridized with nitrocellulose strips containing the following plasmid probes: the first three EcoRI fragments of the PGHSmu-Z cDNA (PGHSZ-l.15, PGHSZ-1.10, PGHSZ-l.65); the PGHSm-l cDNA (PGHS-l), Bz-macroglobulin (B2M); the mouse c-fos (FOS); and the plasmid pUC19 (PUC). (Panel A, 3T3 cells stimulated with 16% fetal calf serum; Panel B, 3T3 cells stimulated with 16% fetal calf serum and 10° M dexamethasone.) Autoradiograms were quantitated by densitometry on a Bioimage Visage 110 image analyzer. Values for the PGHSmu-Z (panels C and D), PGHSmu-l (panel E), and c-fos (panel F) are reported as the integrated optical density of their respective bands, normalized to integrated optical density of the BZ-macroglobulin band. (Panel C, PGHSm-Z transcription in cells stimulated with 16% fetal calf serum; panel D, PGHSm-Z transcription in cells stimulated with 16% fetal calf serum and 106 M dexamethasone; panel B, PGHSm-l transcription cells stimulated with serum or serum plus 106 M dexamethasone; panel F, c-fos transcription in cells stimulated with serum or serum plus 10" M dexamethasone). A single optimal autoradiogram exposure was employed to compare hybridizations of individual plasmid probes between treatment groups, but because of the differential hybridization efficiency, the optimal exposure time varied for each plasmid. PGH82~L10- PGHSZ-L15- PGHSZ-LGS- PGHSfi 82M- . FOS- PUC- PGHS2-L10- PGHSZ-L15- ‘ PGHSZ-LGS- PGHSL 82M- FOS- PUC- 122 SERUM 0.5 1 2 3 4 6 HR II. 1.. l_- .m- . — — — - .— ...l - - - - .- .. o . .,0 c Q . C O - SERUM+DEX (1 uM) 0.5 1 2 3 4 6 HR ‘__‘. -‘-— FIGURE 35 9 Relative PGHS-2 Transcription 30 N O (PGHSZ/BZM) .5 O .6 Relative PGHS-2 Transcription 30 N O (Pensz/ezu) .L O 123 —D— PGH82-1 .15 —.— PGHSZ-1 .1 + PGHS2-1 .65 l ‘ d T I V Y I 3 4 Time(hr) —U— PGHSZ-1.15 + PGHSZ-i .1 + PGH82-1.65 3 4 Time(hr) d 7 FIGURE 35 124 o . a . .w . r . 1 Eugmzea. E 53:85: 5:9. 2:22. 4 3 Time(hr) FIGURE 35 Time(hr) 1 33.8... cozntomcaah «on. 02.23.. 125 as transcription of the gene decreases indicates that the PGH synthase-2 mRNA has a relatively short half-life, a common feature of many immediate-early genes (96). Although hybridization to the three PGH synthase-2 probes varied, the intensities are about proportional to the sizes of the individual fragments; moreover, the overall pattern of hybridization for each of the PGH synthase—2 plasmids is identical. This indicates that transcription is uniform along the gene and suggests that mRNA synthesis is regulated at the level of initiation of transcription and not elongation. Glucocorticoids have previously been shown to inhibit expression of PGH synthase-2 mRNA in TPA- and growth factor-stimulated cells (54). We have compared the relative effects of dexamethasone on the serum-stimulated expression of PGH synthases-1 and -2; we also examined the effects of dexamethasone on the rates of transcription of these two genes. As previously reported for TPA- and growth factor-stimulated expression (54), serum-stimulated PGH synthase-2 expression is greatly reduced by dexamethasone (Figure 34D,F). The 1C5o for inhibition by dexamethasone was about 3 x 109 M (Figure 36); the value is close to the Kd of 2 x 10'9 M for dexamethasone binding to glucocorticoid receptor (GR) (97). In contrast to previous studies which showed that dexamethasone completely inhibited TPA-stimulated PGH synthase-2 expression (54), we found a maximum of 70% inhibition of peak serum- stimulated PGH synthase-2 expression, even with 10‘5 M dexamethasone (Figure 36). Using nuclear run-off assays, we determined that dexamethasone (10'6 M) reduced serum-stimulated transcription of the PGH synthase-2 gene to the same extent as mRNA expression, about 70% (Figure 35). Again, changes in transcription, as measured by hybridization to the three PGH synthase-2 probes, appear to be uniform along the gene, 126 Figure 36. Dose-response curve for dexamethasone inhibition of the serum-induced PGh synthase-2 mRN A expression. Quiescent 3T3 cells were stimulated with 16% fetal calf serum in the presence of the indicated concentration of dexamethasone. At 1 hour, cells were collected and the mRNA was isolated. (Panel A) Northern blot analysis of 1 hour PGHSmu-Z mRN A expression with increasing dexamethasone concentrations. (Panel B) Relative PGHSmu-Z mRNA expression was determined by densitometry as described in Figure 34, and is plotted as percent expression in the absence of glucocorticoids. Dexamethasone has an IDso value of about 3 x 10'9 M for inhibition of PGHSm-Z mRNA expression. 127 3...: Soé: at... 3.7.: 3.7.: 3°7cw £30m oz PGHS2- GAPDH- . .m. . FIGURE 36 128 — q q ‘1 _ d u H — d u u 1 A 0 o o o 8 6 4 2 ca... .o s. .26.. o._ PGH synthase-2 . PGH synthase-2 90M transcription mRNA translation protein (-) (degradation) Dexamethasone Found In: macrophages ovary libroblasts tracheal epithelial cells prostate HNEC Usually lound ln stimulated tlssues FIGURE 40 150 instability of message probably limits PGH synthase protein expression. Once the PGH synthase protein is formed, it can be autoinactivated during formation of prostaglandins as is PGH synthase-1. PGH synthase-2, unlike PGH synthase-1, is also fairly unstable. Expression in serum-stimulated 3T3 cells and in hCG stimulated rat ovary is transitory, lasting for 3 to 7 hours. The cyclooxygenase activity of PGH synthase-2 is sensitive to most of the same NSAIDS as is PGHSmu-l. Most commonly used N SAIDs preferentially inhibit PGH synthase-l, but we found that one drug, the active metabolite of N abumetone, 6-methoxy-2-naphthyl acetic acid (6-MNA), preferentially inhibits PGH synthase-2. Aspirin does not inhibit the oxygenase activity of PGH synthase-2, but rather causes a shift in the product formation to 15-HETE. Aspirin, which probably acetylates PGH synthase-2, does not block arachidonic acid binding, but does interfere with oxygen addition at carbons 9 and 11. Our model (Figure 40) outlines both the regulation of the two PGH synthases and proposes a reason for two similar enzymes catalyzing the same reaction. PGH synthase-l is responsible for constitutive production of prostaglandins, while PGH synthase-2 is highly inducible and causes the synthesis of high levels of prostaglandins as needed. The differential expression and regulation of these two proteins allows each of them of play unique roles which allow prostaglandin production to be fine-tuned in the body. BIBLIOGRAPHY BIBLIOGRAPHY HEDIN, L., D. GADDY-KURTEN, R. KURTEN, D. L. DEWITI‘, W. L. SMITH, AND J. S. RICHARDS. Prostaglandin endoperoxide synthetase in rat ovarian follicles: content, cellular distribution and evidence for hormonal induction preceding ovulation. Endocrinology 121: 722-731,]987. OATES, J. A., G. A. FITZGERALD, R. A. BRANCH, E. K. JACKSON, H. R. KNAPP, AND L. J. ROBERTS. Clinical implications of prostaglandin and thromboxane A2 formation. N. Engl. J. Med. 319: 689-698,]988. NOORT, W. A., F. A. DEZWART, AND M. J. N. C. KEIRSE. Changes in urinary 6-keto-prostaglandin Fml excretion during pregnancy and labor. Prostaglandins 35: 573-582,]988. OATES, J. A., G. A. FITZGERALD, R. A. BRANCH, E. K. JACKSON, H. R. KNAPP, AND L. J. ROBERTS. Clinical implications of prostaglandin and thromboxane A2 formation: second of two parts. N. Engl. J. Med. 319: 761-767,]988. SMITH, W. L., AND P. BORGEAT. The eicosanoids: prostaglandins, thromboxanes, leukotrienes, and hydroxy-eicosaenoic acids. In: Biochemistry of Lipids and Membranes, first ed., edited by D. E. Vance, and Vance, J. E. Menlo Park: Benjamin/Cummings Publishing Company, Inc., 1985, p. 325-360. SMITH, W. L., AND L. J. MARNE'I'I‘. Prostaglandin endoperoxide synthase: structure and catalysis. BBA 1083: 1-17,1991. MONCADA, S., R. GRYGLEWSKI, AND J. R. VANE. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature 263: 663-665,1976. HAMBERG, M., J. SVENSSON, AND B. SAMUELSSON. Thromboxanes: a new group of biologically active compounds derived from prostaglandin endoperoxides. Proc. Natl. Acad. Sci. USA 72: 2994-2998,]975. 151 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 152 THE STEERING COMMITTEE OF THE PHYSICIANS HEALTH STUDY RESEARCH GROUP. Preliminary report: findings from the aspirin component of the ongoing physicians health study. N. Engl. J. Med. 318: 262-264,]988. BURCH, J. W., N. L. BAENZIGER, N. STANFORD, AND P. W. MAJERUS. Sensitivity of fatty acid cyclooxygenase from human aorta to acetylation by aspirin. Proc. Natl. Acad. Sci. USA 75: 5181-5184,]978. CODDE, J. R, AND L. J. BEILIN. Prostaglandins and experimental hypertension: a review with special emphasis on the effect of dietary lipids. J. Hyperten. 4: 675-686,]986. CRUNKHORN, R, AND A. L. WILLIS. Actions and interactions of prostaglandins administered intradermally in rat and man. Br. J. Pharmacol. 36: 216P-217P,1969. MONCADA, S., S. H. FERREIRA, AND J. R. VANE. Prostaglandins, aspirin-like drugs and the oedema of inflammation. Nature 246: 217-219,1973. ESPEY, L. L. Ovulation as an inflammatory reaction-a hypothesis. Biol. Reprod. 22: 73-106,1980. BAUMINGER, 8., AND H. R. LINDNER. Periovulatory changes in ovarian prostaglandin formation and their hormonal control in the rat. Prostaglandins 9: 737-75 1,1975. WONG, W. Y., D. L. DEWI’I'I‘, W. L. SMITH, AND J. S. RICHARDS. Rapid induction of prostaglandin endoperoxide synthase in rat preovulatory follicles by luteinizing hormone and cAMP is blocked by inhibitors of transcription and u‘anslation. Mol. Endocrinol. 3: l7l4-l723,1989. EGAN, R. W., J. PAXTON, AND F. A. KUEl-IL. Mechansim for irreversable self-deactivation of prostaglandin synthetase. J. Biol. Chem. 251: 7325-7335,l976. HEMLER, M. E., AND W. E. M. LANDS. Evidence for a peroxide-initiated free radical mechanism of prostaglandin biosynthesis. J. Biol. Chem. 255: 6253-6261,1980. RAZ, A., A. WYCHE, N. SIEGEL, AND P. NEEDLEMAN. Regulation of fibroblast cyclooxygenase synthesis by interleukin-1. J. Biol. Chem. 263: BUZZ-3028,1988. CASEY, M. L., K. KORTE, AND P. C. MACDONALD. Epidermal growth factor stimulation of prostaglandin E, biosynthesis in amnion cells. J. Biol. Chem. 263: 7846-7854,]988. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 153 HABENICHT, A. J. R., M. GOERIG, J. GRULICH, D. ROTHE, R. GRONWALD, U. LOTH, G. SCHETTLER, B. KOMMERELL, AND R. ROSS. Human platelet-derived growth factor stimulates prostaglandin synthesis by activation and by rapid de novo synthesis of cyclooxygenase. J. Clin. Invest. 75: 1381-1387,1985. WU, K. K., H. HA'I'ZAKIS, S. 8. L0, D. C. SEONG, S. K. SANDUJA, AND H. TAI. Stimulation of de novo synthesis of prostaglandin G/H synthase in human endothelial cells by phorbol ester. J. Biol. Chem. 263: 19043-19047,l988. DEWI'IT, D. L. Prostaglandin endoperoxide synthase: regulation of enzyme expression. BBA 1083: 121-134,1991. LIN, A. H., M. J. BIENKOWSKI, AND R. R. GORMAN. Regulation of prostaglandin H synthase mRNA levels and prostaglandin biosynthesis by platelet-derived growth factor. J. Biol. Chem. 264: l7379-17383,l989. DEWI'IT, D. L., AND E. A. MEADE. Serum and glucocorticoid regulation of gene transcription and expression of the prostaglandin synthase-1 and prostaglandin synthase-2 isozymes. J. Biol. Chem. submitted,l992. KUJUBU, D. A., B. S. FLETCHER, B. C. VARNUM, R. W. LIM, AND H. R. HERSCHMAN. TISlO, a phorbol ester tumor promoter inducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthase/cyclooxygenase homologue. J. Biol. Chem. 266: 12866-12872,l991. SIMONSON, M. S., J. A. WOLFE, M. KONIECZKOWSKI, J. R SEDOR, AND M. J. DUNN. Regulation of prostaglandin endOperoxide synthase gene expression in cultured glomerular rat mesangial cells: activation by serum. In: Prostaglandins and Related Compounds 1], edited by B. Samuelsson. New York: Raven Press, Vol. 21, 1990, p. in press. SIMMONS, D. L., D. B. LEVY, Y. YANNONI, AND R. L. ERIKSON. Identification of a phorbol ester-repressible v-src-inducible gene. Proc. Natl. Acad. Sci. USA 86: 1178-1182,]989. O’BANION, M. K., H. B. SADOWSKI, V. WINN, AND D. A. YOUNG. A serum- and glucocorticoid-regulated 4 kb mRN A encodes a cyclooxygenase-related protein. J. Biol. Chem. 266: 23261-23267,]991. O’BANION, M. K., V. D. WINN, AND D. A. YOUNG. cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc. Natl. Acad. Sci. USA 89: 4888-4892,]992. 31. 32. 33. 34. 35. 36. 37. 38. 39. 154 BURCH, R. M., J. R. CONNOR, AND J. AXELROD. Interleukin 1 amplifies receptor-mediated activation of phospholipase A2 in 3T3 fibroblasts. Proc. Natl. Acad. Sci. USA 85: 6306-6309,l988. RAZ, A., A. WYCHE, AND P. NEEDLEMAN. Temporal and pharmacological division of fibroblast cyclooxygenase expression into transcriptional and translational phases. Proc. Natl. Acad. Sci. USA 86: 1657-1661,]989. RAZ, A., A. WYCHE, J. FU, K. SEIBERT, AND P. NEEDLEMAN. Regulation of prostanoid synthesis in human fibroblasts and human blood monocytes by interleukin- 1 , endotoxin, and glucocoticoids. In: Advances in Prostaglandin, Thromboxane, and Lekotriene Research, edited by B. Samuelsson. New York: Raven Press, Ltd, Vol. 20, 1990, p. 22-27. ROSSI, V., F. BREVIARIO, P. GHEZZI, E. DEJANA, AND A. MANTOVANI. Prostacyclin synthesis induced in vascular cells by interleukin-1. Science 229: 174-176,1986. MAIER, J. A., T. HLA, AND T. MACIAG. Cyclooxygenase is an immediate-early gene induced by interleukin-1 in human endothelial cells. J. Biol. Chem. 265: 10805-10808,1990. ZAKAR, T., AND D. M. OLSON. Stimulation of human amnion prostaglandin E1 production by activators of protein kinase C. J. Clin. Endo. Metab. 67: 915-923,]988. GOERIG, M., A. J. HABENICHT, R. HEITZ, W. ZEH, H. KATUS, B. KOMMERELL, R. ZIEGLER, AND J. A. GLOMSET. sn-l,2-Diacyglycerols and phorbol diester stimulate thromboxane synthesis by de novo synthesis of prostaglandin H synthase in human promyelocytic leukemia cells. J. Clin. Invest. 79: 903-907,1987. SIROIS, 1., AND J. S. RICHARDS. Purification and characterization of a novel, distinct isoforrn of prostaglandin endoperoxide synthase induced by human chorionic gonadotropin in granulosa cells of rat preovulatory follicles. J. Biol. Chem. 267: 6382—6388,l992. SIROIS, J., D. L. SIMMONS, AND J. S. RICHARDS. Hormonal regulation of messenger ribonucleic acid encoding a novel isoform of prostaglandin endoperoxide H synthase in rat preovulatory follicles. J. Biol. Chem. 267: 11586-11592,]992. 40. 41. 42. 43. 45. 46. 47. 48. 155 EGGLESTON, D. L., C. WILKEN, E. A. VAN KIRK, R. G. SLAUGHTER, T. H. JI, AND W. J. MURDOCH. Progesterone induces expression of endometrial messenger RNA encoding for cyclooxygenase (sheep). Prostaglandins 39: 675-683,l990. HUSLIG, R. L., R. L. FOGWELL, AND W. L. SMITH. The prostaglandin forming cyclooxygenase of ovine uterus: relationshp to luteal function. Biol. Reprod. 21: 589-600,1979. CASEY, M. L., M. D. MITCHELL, AND P. C. MACDONALD. Epidermal growth factor-stimulated prostaglandin E, production in human amnion cells: specificity and nonesterified arachidonic acid dependency. Mol. Cell. Endocrinol. 53: 169-176,1987. ZAKAR, T., AND D. M. OLSON. The action of epidermal growth factor on human amnion prostaglandin E2 output. Can. J. Physiol. Pharmacol. 66: 769-775,]988. YOKOTA, K., M. KUSAKA, T. OHSHIMA, S. YAMAMOTO, N. KURIHARA, T. YOSHINO, AND M. KUMEGAWA. Stimulation of prostaglandin E2 synthesis in cloned osteoblastic cells of mouse (MC3T3-E1) by epidermal growth factor. J. Biol. Chem. 261: 15410-15415,1986. BAILEY, J. M., B. MUZA, T. HLA, AND K. SALATA. Restoration of prostacyclin synthase in vascular smooth muscle cells after aspirin treatment: regulation by epidermal growth factor. J. Lipid Res. 26: 54-61,1985. FRASIER-SCO'I'I‘, K., H. HATZAKIS, D. SEONG, C. M. JONES, AND K. K. WU. Influence of natural and recombinant interleukin-2 on endothelial cell arachidonate metabolism. J. Clin. Invest. 82: 1877-1883,]988. YAMAMOTO, S., T. OSHIMA, K. MASAMI, K. YOKOTA, T. YOSHIMOTO. K. SUMITANI, T. KAWATA, AND M. KUMEGAWA. Molecular mechanism of cyclooxygenase regulation in cloned mouse osteoblastic cell MC3T3—E1. In: Advances in Prostaglandin, Thromboxane, and Leukotriene Research, Vol. 19 ed., edited by B. Samuelsson, Wong, Y.-K., and Sun, S. S. New York: Raven Press, Ltd. New York, Vol. 19, 1989, p. 444-449. BURCH, R. M., A. LUINI, D. E. MAIS, D. CORDA, J. Y. VANDERHOEK, L. D. KOHN, AND J. AXELROD. aradrenergic stimulation of arachidonic acid release and meatabolism in a rat thyroid cell line. J. Biol. Chem. 261: 11236-11241,1986. 49. 50. 51. 52. 53. 54. 55. 56. 57. 156 BURCH, R. M., A. LUINI, AND J. AXELROD. Phospholipase A2 and phospholipase C are activated by distinct GTP-binding proteins in response to a,-adrenergic stimulation in FRTL5 thyroid cells. Proc. Natl. Acad. Sci. USA 83: 7201-7205,1986. SUMITAN I, K., K. TERUSHIGE, T. YOSHIMOTO, S. YAMAMOTO, AND M. KUMEGAWA. Fatty acid cyclooxygenase activity stimulated by transforming growth factor-B in mouse osteoblastic cells (MC3T3-El ). Arch. Biochem. Biophys. 270: 588-595,]989. BIENKOWSKI, M. J., M. A. PETRO, AND L. J. ROBINSON. Inhibition of thromboxane A2 synthesis in U937 cells by glucocorticoids. J. Biol. Chem. 264: 6536-6544,1989. MASFERRER, J. L., B. S. ZWEIFEL, K. SEIBERT, AND P. NEEDLEMAN. Selective regulation of cellular cyclooxygenase by dexamethasone and endotoxin in mice. J. Clin. Invest. 86: 1375-1379,]990. FU, J., J. L. MASFERRER, K. SEIBERT, A. RAZ, AND P. NEEDLEMAN. The induction and suppression of prostaglandin H2 synthase (cyclooxygenase) in human monocytes. J. Biol. Chem. 265: l6737-l6740,1990. KUJUBU, D. A., AND H. R. HERSCHMAN. Dexamethasone inhibits rnitogen induction of the T18 10 prostaglandin synthase/cyclooxygenase gene. J. Biol. Chem. 267: 7991-7994,l992. GOERIG, M., A. J. R. HABENICHT, W. ZEH, P. SALBACH, B. KOMMERELL, D. E. R. ROTHE, W. NASTAINCZYK, AND J. A. GLOMSET. Evidence for coordinate, selective regulation of eicosanoid synthesis in platelet-derived growth factor-stimulated 3T3 fibroblasts and in HL-60 cells induced to differentiate into macrophages or neutrophils. J. Biol. Chem. 263(36 Suppl. 25 Dec): 19384-19391,1988. HABENICHT, A. J. R., H. A. DRESEL, M. GOERIG, JOERG-A. WEBER, M. STOEHR, J. A. GLOMSET, R. ROSS, AND G. SCHETTLER. Low density lipoprotein receptor-dependent prostaglandin synthesis in Swiss 3T3 cells stimulated by platelet-derived growth factor. Proc. Natl. Acad. Sci. USA 83: 1344-1348,1986. KOEHLER, L., R. HASS, K. WESSEL, D. L. DEWI'IT, V. KAEVER, K. RESCH, AND M. GOPPELT-STRUEBE. Altered arachidonic acid metabolism during differentiation of the human monoblastoid cell line U937. Bioch. Biophys. Acta 1042: 395-403,l990. ‘58. 59. 61. 62. 63. 65. 66. 67. 157 BROOKS, P. M., AND R. 0. DAY. Nonsteroidal antiinflammatory drugs- differences and similarities. N. Engl. J. Med. 324: 17l6-1725,1991. FLOWER, R. J ., S. MONCADA, AND J. R. VANE. Analgesic-antipyretics and anti-inflammatory agents; drugs employed in the treatment of gout. In: Goodman and Gilman’s The Pharmacological Basis of Therapeutics, seventh ed., edited by A. G. Gilman, Goodman, L. S., Rall, T. W., and Murad, F. New York: Macmillan Publishing Company, 1985, p. 674-715. DEWITI‘, D. L., E. A. MEADE, E. A. EL-HARITH, AND W. L. SMITH. Prostaglandin endoperoxide G/H synthase as a regulatory enzyme in prostaglandin biosynthesis. In: Platelets and Vascular Occlusion, edited by C. Patrono, and FitzGerald, G. A. New York: Raven Press, Vol. 54, 1989, p. 109-118. KRAEMER, S. A., E. A. MEADE, AND D. L. DEWI'IT. Prostaglandin endoperoxide synthase gene structure: identification of the transcriptional start site and 5’-flanking regulatory sequences. Arch. Biochem. Biophys. 293: 391-400,l992. WONG, W. Y. L., AND J. S. RICHARDS. Evidence for two antigenically distinct molecular weight variants of prostaglandin H synthase in the rat ovary. Mol. Endocrinol. 5: 1269-1279,199l. ROSEN, G. D., T. M. BIRKENMEIER, A. RAZ, AND M. J. HOLTZMAN. Identification of a cylcooxygenase-related gene and its potential role in prostaglandin formation. Biochem. Biophys. Res. Commun. 164: 1358-1365,1989. XIE, W., J. G. CHIPMAN, D. L. ROBERTSON, R. L. ERIKSON. AND D. L. SIMMONS. Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc. Natl. Acad. Sci. USA 88: 2692-2696, 1991 . ROME, L. H., AND W. E. M. LANDS. Structural requirements for time-dependent inhibition of prostaglandin biosynthesis by anti-inflammatory drugs. Proc. Natl. Acad. Sci. USA 72: 4863-4865,]975. DEWI'I'I‘, D. L., E. A. EL-HARITH, S. A. KRAEMER, M. J. ANDREWS, E. F. YAO, R. L. ARMSTRONG, AND W. L. SMITH. The aspirin and heme-binding sites of ovine and murine prostaglandin endoperoxide synthases. J. Biol. Chem. 265: 5192-5198,1990. MONCADA, 8., AND J. R. VANE. Pharmacology and endogenous roles of prostaglandin endoperoxides, thromboxane A2 and prostacyclin. Pharmacol. Rev. 30: 293-331,1979. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 158 SAMBROOK, J., E. F. FRITSCH, AND T. MANIATIS. Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1989. DEWITI', D. L., AND W. L. SMITH. Primary structure of prostaglandin G/H synthase from sheep vesicular gland determined from the complementary DNA sequence. Proc. Natl. Acad. Sci. USA 85: 1212-1416,1988. FEINBURG, A. R, AND B. VOGELSTEIN. A technique for radioloabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 137: 266-267,1984. SHAW, G., AND R. KAMEN. A conserved AU sequence from the 3’ untranslated region of GM-CSF mRNA mediates selective degradation. Cell 46: 659-667,]986. T-vector pUC19 is cut with Smal which leaves blunt ends, then treated with Taq polymerase to thymidylate the blunt cut ends. These vectors increase the efficiency of sublconing of PCR-generated fragments because Taq polymerase often adds a single extra adenine to PCR fragments during the amplification process. MARCHUK, D., M. DRUMM, A. SAULINO, AND F. S. COLLINS. Construction of T-vectors, a rapid and general system for direct cloning of unmodified PCR products. Nucl. Acids Res. 19: 1154,1991. SHIMOKAWA, T., AND W. L. SMITH. Essential histidines of prostaglandin endoperoxide synthase. J. Biol. Chem. 266: 6168-6173,1991. NEEDLEMAN, S. B., AND C. D. WUNSCH. A general method applicable to the search for similarities in the amino acid sequence of two proteins. J. Mol. Biol. 48: 443-453,]970. SMITH, T. E, AND M. S. WATERMAN. Comparison of biosequences. Adv. Appl. Math. 2: 482-482,]981. DEVEREUX, J., P. HAEBERLI, AND O. SMITHIES. A comprehensive set of sequence analysis programs for the VAX. Nucl. Acids Res. 12: 387-395,]984. DAYHOFF, M. O., R. M. SCHWARTZ, AND B. C. ORCUTT. A model of evolutionary change in proteins. In: Atlas of Protein Sequence and Structure, edited by M. O. Dayhoff. Washington, DC: National Biomedical Research Fondation, Vol. 5, suppl. 3, 1978, p. 345-358. OHTA, A., M. C. MAYO, N. KRAMER, AND W. E. M. LANDS. Rapid analysis of fatty acids in plasma lipids. Lipids 25: 742-747,]990. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 159 KROMHOUT, D., E. B. BOSSCHIETER, AND C. COULANDER. The inverse relationship between fish consumption and 20-year mortality from coronary heart disease. N. Engl. J. Med. 312: 1205-1209,]985. FUNK, C. D., L. B. FUNK, M. E. KENNEDY, A. S. PONG, AND G. A. FITZGERALD. Human platelet/erythroleukemia cell prostaglandin G/H synthase: cDNA cloning, expression, and gene chromosomal assignment. FASEB 5: 2304-2312,1991. MANN, C. C., AND M. L. PLUMMER. The Aspirin Wars, New York: Alfred A. Knopf, 1991. FLOWER, R. J. Drugs which inhibit prostaglandin biosynthesis. Pharmacol. Rev. 26: 33-67,1974. SIMMONS, D. L., W. XIE, J. G. CHIPMAN, AND G. E. EVET'I‘. Multiple cyclooxygenases: cloning of a mitogen-inducible form. In: Prostaglandins, Leukotrienes, Lipoxins, and PAP, edited by J. M. Bailey. New York: Plenum Press, 1991, p. 67—78. SMITH, W. L., D. L. DEWI‘I'I', T. SHIMOKAWA, S. A. KRAEMER, AND E. A. MEADE. Molecular basis for the inhibition of prostanoid biosynthesis by nonsteroidal anti-inflammatory agents. Stroke 21: IV24-IV28,1990. SMITH, W. L., AND W. E. M. LANDS. Oxygenation of polyunsaturated fatty acids during prostaglandin biosynthesis by sheep vesicular glands. Biochem. 11: 3276-3282,]972. TURK, J., R. L. MAAS, A. R. BRASH, L. J. ROBERTS, AND J. A. OATES. Arachidonic acid lS-lipoxygenase products from human eosinophils. J. Biol. Chem. 257: 7068-7076,1982. MASFERRER, J. L., K. SEIBERT, B. ZWEIFEL, AND P. NEEDLEMAN. Endogenous glucocorticoids regulate an inducible cyclooxygenase enzyme. Proc. Natl. Acad. Sci. USA 89: 3917-3921,1992. O’SULLIVAN, G. M., F. H. CHILTON, E. M. HUGGINS JR., AND C. E. MCCALL. Lipopolysaccharide priming of alveolar macrophages for enhanced synthesis of prostanoids involves induction of a novel prostaglandin H synthase. J. Biol. Chem. in press,1992. VANDERHOEK, J. Y., R. W. BRYANT, AND J. M. BAILEY. Inhibition of leukotriene biosynthesis by the leukocyte product 15-hydroxy-5,8,11,13-eicosatetraenoic acid. J. Biol. Chem. 255: 10064-10066,l980. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 160 GOETZL, E. J. Selective feed-back inhibition of the 5-lipoxygenation of arachidonic acid in human T-lymphocytes. Biochem. Biophys. Res. Commun. 101: 344-350,]981. SHIMOKAWA, T., R. J. KULMACZ, D. L. DEWI'IT, AND W. L. SMITH. Tyrosine 385 of prostaglandin endoperoxide synthase is required for cyclooxygenase catalysis. J. Biol. Chem. 265: 20073-20076,1990. YOKOYAMA, C., AND T. TANABE. Cloning of the human gene encoding prostaglandin endoperoxide synthase and primary su‘ucture of the enzyme. Biochem. Biophys. Res. Commun. 165: 888-894,1989. FLETCHER, B. S., D. A. KUJUBU, D. M. PERRIN, AND H. R. HERSCHMAN. Structure of the mitogen-inducible TISlO gene and demonstration that the T18 lO-encoded protein is a functional prostaglandin G/H synthase. J. Biol. Chem. 267: 4338-4344,1992. BEDARD, PIERRE-A., D. ALCORTA, D. L. SIMMONS, KA-C. LUK, AND R. L. ERIKSON. Constitutive expression of a gene encoding a ppolypeptide homologous to biologically active human platelet protein in Rous sarcoma virus-transformed fibroblasts. Proc. Natl. Acad. Sci. USA 84: 6715-6719,]987. HERSCHMAN, H. R. Primary response genes induced by growth factors and tumor promoters. In: Annual Review of Biochemistry, edited by C. C. Richardson, Abelson, J. N., Meister, A., and Walsh, C. T. W. Palo Alto: Annual Reviews Inc., Vol. 60, 1991, p. 281-319. BRANDON, D. D., A. J. MARKWICK. B. P. CHROUSOS, AND D. L. LORIAUX. Glucocorticoid resistance in humans and nonhuman primates. Cancer Res. 49: 2203-2213,l989. KAWAHARA, R. S., ZHENG-W. DENG, AND T. F. DEUEL. Glucocorticoids inhibit the transcriptional induction of IE, a platelet-derived growth factor-inducible gene. J. Biol. Chem. 266: l3261-13266,1991. JONAT, C., H. J. RAHMSDORF, K. PARK, A. C. B. CATO, S. GEBEL, H. PONTA, AND P. HERRLICH. Antitumor promotion and antiinflammation: down-modulation of AP—l (Fos/Jun) activity by glucocorticoid hormone. Cell 62: 1189- 1204,1990. AKERBLOM, I. E., E. P. SLATER, M. BEATO, J. D. BAXTER, AND P. L. MELLON. Negative regulation by glucocorticoids through interference with a CAMP responsive enhancer. Science 241: 350-353,1988. 101. 102. 103. 161 SAKAI, D., S. HELMS, J. CARLSTEDT—DUKE, JAN-A. GUSTAFSSON, F. M. ROTTMAN, AND K. R. YAMAMOTO. Hormone-mediated repression: a negative glucocorticoid response element from the bovine prolactin gene. Genes Dev. 2: 1144-1154,1988. DROUIN, J., M. TIFIRO, R. PLANTE, M. NEMER, P. ERIKSSON, AND O. WRANGE. Glucocorticoid receptor binding to a specific DNA seqeunce is required for hormone-dependent repression of pro-opiomelanocortin gene transcription. Mol. Cell. Biol. 9: 5305-5314,1989. OSHIMA, T., T. YOSHIMOTO, S. YAMAMOTO, M. KUMEGAWA, C. YOKOYAMA, AND T. TANABE. CAMP-dependent induction of fatty acid cyclooxygenase mRNA in mouse osteoblastic cells (MC3T3-E1). J. Biol. Chem. 266: 13621-13626,]991.